Chem Soc Rev

View Article Online

REVIEW ARTICLE View Journal | View Issue

Inorganic nanostructures for photoelectrochemical and

Cite this: Chem. Soc. Rev., 2013,
photocatalytic water splitting
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.

42, 2294
Frank E. Osterloh*

Received 16th July 2012
DOI: 10.1039/c2cs35266d
www.rsc.org/csr
The increasing human need for clean and renewable energy has stimulated research in artificial
photosynthesis, and in particular water photoelectrolysis as a pathway to hydrogen fuel. Nanostructured
devices are widely regarded as an opportunity to improve eciency and lower costs, but as a detailed
analysis shows, they also have considerably disadvantages. This article reviews the current state of
research on nanoscale-enhanced photoelectrodes and photocatalysts for the water splitting reaction. The
focus is on transition metal oxides with special emphasis of Fe2O3, but nitrides and chalcogenides, and
main group element compounds, including carbon nitride and silicon, are also covered. The eects of
nanostructuring on carrier generation and collection, multiple exciton generation, and quantum
confinement are also discussed, as well as implications of particle size on surface recombination, on the
size of space charge layers and on the possibility of controlling nanostructure energetics via potential
determining ions. After a summary of electrocatalytic and plasmonic nanostructures, the review concludes
with an outlook on the challenges in solar fuel generation with nanoscale inorganic materials.

Introduction
The goal of keeping global temperature increases below 2 1C
compared to pre-industrial times requires the rapid deployment
Department of Chemistry, University of California, Davis, CA 95616, USA.
E-mail: fosterloh@ucdavis.edu; Fax: +1 530 752 8995; Tel: +1 530 754 6242
Part of the solar fuels themed issue.
Frank Osterloh is an inorganic
chemist working on nanostructured
solar energy conversion devices. He
received MA and PhD degrees in
chemistry in 1994 and 1997 from
the Carl von Ossietzky University in
Oldenburg, Germany, after working
under the guidance of the late
Prof. Siegfried Pohl. He then
completed postdoctoral training at
Harvard University with Prof.
Richard H. Holm, after which he
Frank E. Osterloh joined the faculty at the Chemistry
Department at the University of
California, Davis, in 2000. Frank Osterlohs current research interests
are centered on the chemical and physical properties of inorganic
nanomaterials, and on their electro- and photocatalytic functions.
of carbon free energy technology that can compete with the
price of fossil fuels. Access to clean energy is also expected to be
a key factor in sustaining population growth. The solar energy
received on the Earths surface meets current and future
human energy demand.1,2 It can be converted into electricity
with up to 43.5% eciency using photovoltaic cells that
are available today. But these cells are more expensive than
virtually free coal, and electricity is dicult to store and to
distribute over long distances. Such problems are solved
by converting the photochemical energy directly into fuel.
The simplest scheme employs the photoelectrolysis reaction
according to eqn (1) to produce hydrogen from water.
H2O - 12O2(g) + H2(g); DG = +237 kJ mol1 (1)
This hydrogen could become the central energy carrier of a
hydrogen-based economy.2 Alternatively it could be used as an
energy-rich reagent for the exothermic formation of methane,
methanol, or even hydrocarbons using atmospheric CO2 as a
carbon feedstock. Thus, it could generate conventional fuels for
transportation, which currently accounts for B30% of all
human energy consumption. Among several ways to facilitate
reaction (1) (photovoltaic cell plus electrolyzer, thermochemical
methods), the most economical ones employ photoelectrosynthetic
cells (PECs) and suspended photocatalysts.3 Here a semiconductor
is brought into contact with water and photochemical charge
carriers created under illumination electrolyze the water.

2294 Chem. Soc. Rev., 2013, 42, 2294--2320 This journal is c The Royal Society of Chemistry 2013

Review Article
Fig. 1 Basic photoanodecathode (Schottky-type) and photoanode/photo-
cathode configurations (Tandem or Z-scheme type) for photoelectrochemical
water splitting. Reproduced with permission from ref. 4. Copyright 2010, American
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
Chemical Society.
The best performing photoelectrochemical devices known
today are Tandem cells (Fig. 1), in which the necessary potential
for water electrolysis is produced by a combination of two
or more semiconductors connected in series. The advantage
of this Tandem configuration is that semiconductors with
smaller bandgaps can be used, which absorb a greater fraction
of the solar spectrum.
Eciencies between 12.4% (p-/n-GaAs photoanode and a
GaInP2 photocathode)5 and over 18% (for a multijunction cell),6
have been achieved, i.e. about half of the theoretical eciency
limit for these devices (24.4% for a Tandem7,8 and 30% for a
multijunction device6). But the performance cannot be sustained
because the photoelectrode materials are not stable under
operating conditions and undergo photocorrosion. Similar sta-
bility problems exist for most other visible light absorbing IIVI,
IIIV, and group 14 element semiconductors. Currently, there
are three main strategies to develop better devices that are
ecient, stable, and inexpensive. One is to coat conventional
photovoltaic cells with cocatalysts for water splitting or with
protecting layers to inhibit photocorrosion. This has lead to the
champion water splitting devices mentioned above.5,6 It also
produced to the artificial leaf, a triple junction amorphous
silicon cell, capable of photoelectrolysing water with a solar
energy eciency of up to 4.7%.9,10 Another strategy involves
the development of new metal oxide materials that combine
suitable properties (visible bandgap, chemical stability, high
carrier mobilities, long carrier lifetimes) for photoelectro-
chemical water splitting. Such materials can be made by directed
synthesis, sometimes guided by theory,11 or they can be made by
combinatorial approaches, as described by Bruce Parkinson,1214
Eric McFarland,15 and Nathan Lewis.16 The third strategy is to
exploit scaling laws and specific eects at the nanoscale to enhance
the eciency of existing semiconductors and metal oxides. This
nanostructuring strategy has gained significant interest in the last
20 years, and is the focus of this review article.
Brief history of nanoscale
photoelectrochemistry and photocatalysis
As a research area, nanoscale photoelectrochemistry represents
a fusion of photoelectrochemistry and nanoscience. Modern
This journal is c The Royal Society of Chemistry 2013
View Article Online
Chem Soc Rev
nanoscience is only about 40 years old, and began in 1974 with
Dingles discovery of quantum size eects in thin films,17 and
Arnim Henglein18,19 and Louis Brus20 work on semiconductor
particles (quantum dots). Photoelectrochemistry started in
1955 with Brattains electrochemical studies on germanium
electrolyte junctions,21 and then grew significantly through the
works of Gerischer,22 Pleskov,23 Marcus,24 and others. The
discovery of water photoelectrolysis at illuminated TiO2 electro-
des in 1971 by Fujishima and Honda,25,26 established the use
of photoelectrochemical cells for artificial photosynthesis.
After Bards demonstration of the photocatalytic eects of
suspended semiconductor particles,27 Arthur Nozik formulated
the concept of photochemical diodes.28 He assumed that the
basic functions of a photoelectrosynthetic cell (light absorption,
charge separation, water electrolysis) could be copied in
suspended micro- or nanoscale particles that would split water
under illumination. Initially, two basic device configurations
were proposed (additional ones are described in the recent
review by Walter et al.4). One consists of a Schottky type device,
in which a photoanode material is fused to a cathode material
(Fig. 1), and the other configuration is a Tandem or Z-scheme
device, with a photoanode coupled to a photocathode material.
The underlying physical principles of these devices are not
covered here, but have been subject of several reviews.4,2831
The earliest experimental attempts to achieve such photo-
diodes date back to the early eighties, when Gra tzels group
published a series of papers on water splitting with colloidal
structures, such as the one shown in Fig. 2. Even though these
catalysts worked reasonably well in the presence of a chemical
bias (using sacrificial electron donors and acceptors), overall
water splitting with them was not achieved.30,32 Likely, these
catalysts are not able to generate a sucient photoelectrochemical
potential dierence for water electrolysis.
With microscale particles, the situation is slightly better,
and many inorganic composites are known today that can
catalyze water photoelectrolysis when suspensions of them
are exposed to ultraviolet light.3440 Finally, during the last
two years, even four nanoscale photochemical diodes that
function without a chemical bias have been reported in the
literature (Fig. 3). All of them are Schottky type devices, all
require ultraviolet light for operation, but only one of them has
Fig. 2 Photoelectrochemical Diode after Duonghong. Reproduced with permis-
sion from ref. 33. Copyright 1981, American Chemical Society.
Chem. Soc. Rev., 2013, 42, 2294--2320 2295

Chem Soc Rev
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
Fig. 3 The only four known nanoscale photocatalysts for overall water splitting.
Reproduced with permission from ref. 4143 and 45. Copyright 2011, Royal
Society of Chemistry. Copyright 2012, American Chemical Society.
reasonable eciency.41 One system was developed by Akihiko
Kudos group, and consists of LiNbO3 nanowires (70 nm 
10 mm) with a bandgap of 4.0 eV.42 After modification with 1%
(mass) RuO2 cocatalyst, the nanocomposite can split water with
0.7% quantum eciency at 254 nm (O2 was evolved in slight
excess). In a second example, Yan et al. reported overall water
splitting with 3% (mass) RuO2-modified Zn2GeO4 nanorods
(100  150 nm) under UV light from a 400 W Hg UV lamp to
excite the large bandgap (>4.5 eV) of the material. Here, the H2
evolution rate was 17.4 mmol h1 (stoichiometric O2) with
100 mg of the catalyst.43 Last year, Kazunari Domens group
reported overall water splitting with aggregates of NiOx-loaded
NaTaO3 nanocrystals (20 and 40 nm diameter).41 The catalytic
rate of this system was very high (2.0 mmol h1 of H2 with
stoichiometric O2), but it likely requires photons with very high
energy, due to the 4.0 eV band gap of bulk-NaTaO3.44 The last
known nanoscale water splitting system consists of 30 nm
SrTiO3 nanocrystals with a 3.3 eV indirect bandgap.45
After modification with a NiOx cocatalyst, this system produces
19.4 mmol H2 g1 h1 (with stoichiometric O2) from pure water
and under >3.2 eV illumination (26.3 mW cm2).
While bias-free photocatalytic water splitting with nanostructure
catalysts remains a significant challenge, a lot of progress has been
made, especially over the past 10 years, in the area of electrically-
biased nanostructured photoelectrodes. The most prominent
examples are based on nanostructured Fe2O3, which in combi-
nation with a nanoscale IrO2 co-catalyst, supports photo-
currents of up to 3.0 mA cm2 under AM 1.5 solar illumination
and with a 1.23 V applied bias.46 This exceeds the photocurrent
of Fe2O3 single crystal electrodes by over two orders of magnitude.47
Nanostructure performance enhancements have also been
observed with WO3, MnO2, BiVO4, and carbon nitride (see
below), and with many cocatalyst materials, including MoS2
and with several noble metals. These devices owe their improved
eciency to an increase in surface area by the nanostructured
materials, and to a shortening of charge transport pathways
2296 Chem. Soc. Rev., 2013, 42, 2294--2320
View Article Online
Review Article
between the material interior and the surface. These advantages
of nanomaterials for solar energy conversion were first demon-
strated with Gra tzels nanostructured dye-sensitized photo-
electrochemical cell.48 In addition, that device also showed that
nanomaterials can be used to interface molecular and bulk
components,49,50 allowing to bridge the gap between the dier-
ent length scales. Unfortunately, the Gra tzel cell design requires
considerable adjustment and optimization before it can be
applied to photoelectrochemical water splitting. The main issue
is that the water splitting reaction is significantly more endergonic,
mechanistically complex, kinetically slower, and chemically more
corrosive than the electron transfer-only processes in regenerative
photoelectrochemical cells. This places exceptional demands on
the stability of the components and on their performance. The
addition of two catalysis functions (for water reduction and
oxidation) to the photovoltaic one (light absorption, charge separa-
tion, charge transport), also increases device complexity. Most of
these functions are connected, and jointly depend on the electrical,
chemical, and optical properties of the components. Adjusting the
nanoscale morphology of the photoelectrode can improve some of
these electrode functions, but at the same it diminishes others.
Therefore, a systematic optimization of photoelectrodes requires a
whole system approach that considers both positive and negative
eects of nanoscaling. These advantages and disadvantages are
discussed in the following.
Pros and cons of nanostructured
photoelectrodes and photocatalysts
1. Shortened carrier collection pathways. Photoexcitation produces
charge carriers with finite mobility and lifetime, depending on
the material, the carrier type, and the light intensity. To drive
water redox reactions, these carriers need to reach the materials
interfaces at the electrolyte and at the back contact. In the
absence of an external field, charge carriers move by diusion
and their range is defined by the mean free diusion length L.
The parameter L depends on the carrier diusion constant D and
the carrier lifetime t (eqn (2)), and a dimensionality factor (q =
2,4,6 for one-, two-, or three-dimensional diusion).
L2 qDt (2)
For intrinsic semiconductors, usually Le > Lh because of the
larger diusion constant D of the electrons compared to holes.
For example, Si has De = 49 cm2 s1 and Dh = 13 cm2 s1
(calculated from mobilities, me = 1900 cm2 V1 s1 and mh =
500 cm2 V1 s1 at 298 K using EinsteinSmoluchowski relation).51
Assuming te = th = 106 s, Le = 98 mm and Lh = 51 mm for one-
dimensional diusion (q = 2). Upon doping, the concentration of
the majority carriers increases, and with it their t and L values. On
the other hand, the lifetime and diusion length of the minority
carriers decrease. For optimum collection of both carrier types at
the back contact, the semiconductor film thickness d has to be in
the same range as Le and Lh (Fig. 4). To improve minority carrier
collection at the scelectrolyte interface, the surface roughness
of the film can be increased, as shown in shown in Fig. 4b.
This journal is c The Royal Society of Chemistry 2013

Review Article
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
Fig. 4 Charge Collection in flat and nanostructured films and in particle
suspensions. d: film or particle thickness; Le: electron diusion length; Lh: hole
diusion length. Suspended particles require both electrons and holes to be
collected at the scelectrolyte interface.
This surface nanostructuring approach is particularly useful for
1st row transition metal oxides (MnO2, Fe2O3), which suer
from low hole mobility and lifetimes.52,53 Ideal electron/hole
collection is possible with suspended nanoparticles, if their
particle size d o Le, Lp. Although in this situation, both carrier
types need to be extracted at the scelectrolyte interface. Thus
there is a need for selective redox agents. This has been referred
to by Helmut Tributsch as kinetic rectification.54
2. Improved light distribution. The ability of a material to
absorb light is determined by the Lambert Beer law and the
wavelength-dependent absorption coecient a. The light pene-
tration depth a1 refers to the distance after which the light
intensity is reduced to 1/e of the original value. For example, for
Fe2O3, a1 = 118 nm at l = 550 nm,7 for CdTe, a1 = 106 nm
(550 nm),55 and for Si, a1 = 680 nm (510 nm).55 To ensure
>90% absorption of the incident light, the film thickness must
be >2.3 times the value of a1 (Fig. 5). Surface-structuring on
the micro- or nanoscale can increase the degree of horizontal
light distribution via light scattering (Fig. 5 and 26). This
trapped light would otherwise be lost by direct reflection from
a flat surface. Light scattering is maximal in particle suspen-
sions, because it can occur at both the front and back sides of
the particles. The eect of light management on the energy
Fig. 5 Light distribution in flat and nanostructured films, and in particle
suspensions. d: film or particle thickness; a1: optical penetration depth. Short
arrows signify scattered or reflected light.
This journal is c The Royal Society of Chemistry 2013
View Article Online
Chem Soc Rev
Fig. 6 Quantum size eect in nanocrystals. After Holmes et al.62
conversion eciency of photovoltaic cells has recently been
discussed by Polman and Atwater,56 see also Fig. 26.
3. Quantum size confinement. The dependence of semiconductor
energetics on particle size has been established in the mid 80s.57
The quantum size eect depends also on the material and the
nanocrystal shape. With increasing the band gap, the conduction
band edge shifts to more reducing and the valence band to more
oxidizing potentials. From MarcusGerischer theory it is expected
that this increase in thermodynamic driving force increases the
rates for interfacial charge transfer and water electrolysis
(Fig. 6).24,58,59 Indeed, this can be experimentally observed for
charge transfer across solidsolid interfaces.60,61 A logarithmic
dependence of the photocatalytic proton reduction rate on the
bandgap has also been verified with CdSe quantum dots.62 These
eects will likely be used more often in advanced solar water
splitting devices.
4. Potential determining ions (PDI). The eect of potential
determining ions on the interfacial energetics is well
known.63,64 For example, the flatband potential of TiO2 is a
linear function of the solution pH.65 Due to the small thickness
of nanostructures, external electric fields can reach into the
nanomaterial interior and modify the local energetic structure.
Thus, the band edge potentials of nanomaterials, and resulting
functions, incl. interfacial charge transfer can be controlled
with PDIs (Fig. 7). Examples are the band shift caused by
hydrosulfide on nano-Bi2S3 in TiO2 nanocrystal films,66 and
the eect of pH on interfacial charge transfer.67
5. Surface area-enhanced charge transfer. The larger specific
surface area of nanomaterials promotes charge transfer across
the material interfaces (solidsolid and solidliquid), allowing
water redox reactions to occur at relatively low current densities
and, correspondingly, low overpotentials. In other words, the
increase of surface area allows to better match the photocurrents
with the slow kinetics of the water redox reactions. In particular
Fig. 7 Eect of PDIs on nanocrystal energetics.
Chem. Soc. Rev., 2013, 42, 2294--2320 2297

Chem Soc Rev
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
Fig. 8 Enhanced photovoltaic eciency in QD solar cells by impact ionization
(inverse Auger eect). This is promoted in QDs due to the lower rates of carrier
relaxation. Reproduced with permission from ref. 72. Copyright 2002, Elsevier
Science Ltd.
water oxidation is known to require milliseconds to seconds to
proceed at Fe2O3 and TiO2, according to recent transient absorp-
tion measurements.68,69 Thus, increases of surface area reduce
the need for highly active, and often expensive cocatalysts, based
on Ir, Rh or Pt.
6. Multiple exciton generation. The altered electronic struc-
ture of strongly size-confined nanocrystals gives rise to multiple
exciton generation (MEG), i.e. the formation of several (n)
electronhole pairs after absorption of one photon with an
energy n times of the band gap of the dot (Fig. 8). The MEG
eect is responsible for the abnormally high eciency of PbSe
QD-sensitized TiO2 photoelectrochemical cells (Parkinson),70
and PbSe photovoltaic cells (Nozik).71 The eect has not yet
been applied to water photoelectrolysis. Future MEG-enhanced
water splitting devices will likely be Tandem or multi-junction
devices, because the individual quantum dots cannot produce a
sucient potential for overall water splitting. This is because
for ecient solar energy conversion, the band gaps of the
relevant dots need to be a fraction of the energy of visible light
photons (E = 1.553.1 eV).
While nanostructuring can improve the light harvesting,
charge transport, kinetic, and energetic parameters of photo-
electrosynthetic cells, it also has significant disadvantages that
can lead to reduced power conversion eciency and lower
durability of devices. Some disadvantages originate from the
inherent properties of nanoparticles, e.g. decreased thermo-
dynamic stability tied to increased surface energy. Others are
specifically tied to photochemical energy conversion. These will
be discussed in the following.
1. Increased surface recombination. Electronhole recombina-
tion is the major loss mechanism in excitonic solar cells and in
photocatalysts. Photogenerated charge carriers recombine
through radiative or nonradiative processes in the bulk phase
of the semiconductor, in the depletion region, or at defects at
the surface (Fig. 9). These processes diminish the steady state
2298 Chem. Soc. Rev., 2013, 42, 2294--2320
View Article Online
Review Article
Fig. 9 Recombination pathways for photoexcited carriers in a semiconductor
PEC.4,74 The arrows signify bulk recombination (Jbr), depletion-region recombina-
tion (Jdr), and surface recombination (Jss). Additional loss mechanisms due to
undesired charge transfer are also shown. Electron tunneling through and over
the barrier produce the current densities (Jt) and (Jet). Electron collection by the
back contact and hole collection by the redox couple (e.g., oxidation of water to
O2) are desired processes shown by thick black arrows. Reproduced with
permission from ref. 74. Copyright 2005, American Chemical Society.
concentrations of usable charge carriers and reduce the rates of
water electrolysis and carrier extraction (thick black arrows in
the figure). In nanostructured semiconductors, surface and
interfacial recombination rates are enhanced due to their larger
specific interfacial areas. It might be possible to control these
losses by reducing surface defects through surface treatments.
For example, for macroscopic silicon films, treatment with HF
can suppress surface recombination.73 Analogous treatments
will have to be developed for nanomaterials.
2. Reduced space charge layer thickness. In nanomaterials,
carrier separation (thick arrows in Fig. 9) is more dicult than
in the bulk, because at average doping concentrations (n0 =
1017 cm3) space charge layers are not eective on the nano-
scale.31,7577 I.e., for spherical nanoparticles, the space charge
layer thickness cannot exceed the radius of the particle
(Fig. 10), which restrains the possible barrier height at the
interface. For example, for 16 nm TiO2 nanocrystals (e = 160) with a
charge carrier concentration of n0 = 1017 cm3, ORegan calculated
a barrier height of 0.3 meV under maximum depletion.78 This
means that in the absence of a strong applied bias, the bands in a
nanoparticle are essentially flat, as shown in Fig. 10.
In the absence of interior electrical forces, the relative
electron/hole injection rates into the electrolyte are governed
by the kinetics of interfacial charge transfer alone.75 This leads
to an increase of undesirable processes (current densities Jt and
Jet in Fig. 9) that essentially short-circuit the device. The
absence of an internal mechanism for charge separation also
increases e/h recombination, as discussed above. This is the reason
why ecient collection of majority carriers in nanocrystalline TiO2
films depends on the presence of iodide as an easily oxidizable
electron donor.78,79 Without iodide, only 4% of electrons can be
harvested at the back contact due to recombination with holes
This journal is c The Royal Society of Chemistry 2013
 View Article Online

Review Article Chem Soc Rev

Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.

Fig. 10 Space charge layers in large and small particles. Reproduced with
permission from ref. 31. Copyright 1994, Springer Berlin/Heidelberg.

on the Ru dye (this corresponds to Jet and Jt in Fig. 9). Further-

more, in nanostructures with quantum size eects (e.g. in many
metal chalcogenides), the e/h pairs are confined to a space
smaller than the Bohr exciton radius. Thus, additional energy
is necessary to separate the charges. This is analogous to organic
photovoltaic cells, where a higher exciton binding energy results Fig. 12 Electron hopping in nanostructured films. Reproduced from ref. 86 by
from the lower dielectric constant of organic polymers.80 permission of the publisher (Taylor & Francis Ltd, http://www.tandf.co.uk/journals).
3. Lower absorbed photon flux. A second consequence of the
increased junction area in nanostructured photoelectrochemical
4. Slow interparticle charge transport. In nanocrystalline films,
cells is lower absorbed flux (if light scattering is neglected). This is
charge carriers move by diusion instead of drift.77,8284 As
shown in Fig. 11 for a flat semiconductor liquid junction and for a
a result, charge transport is much slower than in the bulk,
semiconductor wire array. The latter has a larger roughness, and
increasing the chances for recombination and back reactions.85
thus receives a lower flux per unit area of exposed semiconductor.
If the nanoparticles are not fused together, additional barriers
According to the Shockley diode equation (see below)81 the open
arise from interparticle charge transport, which occurs by thermally
circuit voltage of a solar cell is a logarithmic function of the
activated hopping and by electron tunneling (Fig. 12). It depends
absorbed flux Jphot and of the reverse saturation current J0 of
on the interparticle distance and the electrostatic charging energy
the diode. Since Jphot d J0, the voltage decreases by 0.059 V for
of the donor acceptor nanocrystal couple.86,87 Since nanostructured
every decadic decrease of JPhot (i.e. decadic increase of surface
films can be several hundred nanometers thick, the resistance
roughness). This reduces the thermodynamic driving force
losses from charge transport are a significant factor.
available for water electrolysis.
The problem of charge transport over macroscopic distance
 
kT Jphot can be entirely avoided with nanoscale suspended catalysts. In
VOC ln 1 (3) such catalysts all three metric dimensions are on the nanometer
e J0
scale, and good charge transfer is possible due to the proximity
J0, reverse saturation current of diode; Jphot, photocurrent = of photoanode and cathode. However, the performance in these
photon flux times x irradiated area; VOC, open circuit voltage. photoelectrochemical diodes is limited by ineective charge
separation, as discussed above.

Equivalent circuit diagram

Fig. 11 Increased surface roughness leads to reduced absorbed flux. This
reduces the open circuit voltage, and the thermodynamic driving force for water
electrolysis.
An useful way to illustrate the eects of nanostructuring on
photoelectrosynthetic devices is by using equivalent circuit
diagrams (Fig. 13) of the type commonly employed for photo-
voltaic cells. In the diagram the light-absorbing component
corresponds to a photon-driven current source, and the rectifying
component is shown as a diode in parallel (shunt) to it.
In addition, there are parallel and serial resistances RSH and

This journal is c The Royal Society of Chemistry 2013 Chem. Soc. Rev., 2013, 42, 2294--2320 2299

Chem Soc Rev
Fig. 13 Equivalent circuit diagram for a photovoltaic cell. RSH: Shunt resistance,
RS: serial resistance, ISH: leakage current, IL: generation current.
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
RS. For a photovoltaic cell, the load is an electrical device, and
for a fuel-producing cell, the load is due to water electrolysis.81
In the circuit, ISH, RS, and RSH are dependent upon the physical
size of the solar cell. For two otherwise identical cells, a cell with
twice the surface area of another is expected to have half the shunt
resistance and double leakage current I0 because it has twice the
junction area across which current can leak. It will also have half the
series resistance RS because it has twice the cross-sectional area
through which current can flow. For a nanostructured photoelec-
trochemical cell, the decrease in RSH corresponds to the leakage
currents Jet and JT in Fig. 9. If this current exceeds the generation
current IL, water electrolysis will come to a stop. The decrease of
RS due to the reduction in electronhole transport resistances
can lead to improved performance of a nanostructured device.
However, this benefit will only apply to devices that are nano-
scale in all three dimensions, e.g. suspended nanocatalysts,
which do not have a macroscopic thickness, and thus do not
suer from large serial resistance. Lastly, as the semiconductor
particles approach the nanoscale, the reduction of the space
charge layer eectiveness diminishes the rectifying properties.
This also corresponds to reduction of RSH.
The discussion shows that there are both benefits and problems
associated with nanostructuring photoelectrodes, and that
Fig. 14 Types of Nanostructured TiO2 photoanodes. Adapted with permission from ref. 77, 100 and 106. Copyright 1995, 2008, and 2011 American Chemical
Society.
2300 Chem. Soc. Rev., 2013, 42, 2294--2320
View Article Online
Review Article
optimization of nanostructured PEC devices will require a careful
balance between the opposing eects. This balance of course
depends not only on the morphology and size of the device, but
also on the electrode materials themselves, including their ener-
getics, optical absorption, charge transport properties, and water
redox kinetics. Thus it makes sense to organize the following
presentation on nanostructured photoelectrodes by material type.
Metal oxides with closed shells are discussed at the beginning,
followed by metal oxides with partially filled d-orbitals. Then nitrides
and sulfides are presented and main group element compounds.
Sections on nanostructured cocatalysts and plasmonic nanostruc-
tures are given at the end. The review concludes with a brief outlook
and a discussion of the specific challenges of nanoscale-enhanced
photoelectrochemical devices and catalysts. Selected aspects of
nanoscale photoelectrochemical cells and catalysts have been
reviewed elsewhere.77,8896
Metal oxides closed shell systems
TiO2 and titanates
The discovery of photoanodic water oxidation with TiO225,26
prompted investigations on nanostructured TiO2 photoelectrodes,
which due to their higher surface area were expected to have higher
activity. Fig. 14 summarizes four morphologies that have been
studied. They include particulate films of the type employed in the
Gratzel dye sensitized cell,97 nanorod arrays,98,99 branched nano-
rods,100 and nanotube arrays.101
Direct comparison shows that that under equal experimental
conditions, the photocurrent density from water oxidation
increases in going from particles to rods and to branched rods,
mainly due to improved electron transport to the back-contact
and improved hole transport to the solidliquid interface.100 For
example, TiO2 nanotube arrays have been shown to produce O2
with >90% photon conversion eciency (17.5 mA cm2 at 1.5 V
vs. Ag/AgCl in 1 M KOH).102 But the band gap of TiO2 precludes
This journal is c The Royal Society of Chemistry 2013

Review Article
visible light absorption. Due to their large surface area, TiO2
nanowire arrays can be N-doped with ammonia at relatively low
temperature (500 1C), to increase the visible light response to
IPCE values of up to 4% at 450 nm (+1.23 V RHE), and 18%, after
attachment of a Co-cocatalyst.103 The positive eect of N doping
is enhanced if H2 reduction precedes ammonolysis.104 This is
thought to produce Ti3+ in addition to the substitutional N
defects. Recently, a TiO2 nanotube array was integrated with a
CuTiO nanotube array to achieve overall water photoelectrolysis
with an energy eciency of 0.30% and a photocurrent of up to
0.25 mA cm2 (AM 1.5 illumination).105 The performance of
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
nanoparticulate films is lower because electron transport has to
occur by hopping and tunneling between individual particles.
Annealing of the film improves transport, and the mean electron
free diusion length can reach 10 mm (at an applied bias between
0.15 and 0.1 V vs. Ag/AgCl in 3 M KCl and pH 2).107 For TiO2
nanotubes, LD,e increases to 100 mm due to the continuity of the
lattice.106 And for TiO2 nanorods, the electron diusion coecient
is 200 times higher than in nanoparticulate films.98
Electron transport in nanostructured TiO2 also is very sensi-
tive to the electron lifetime. For example, oxygen generated
during water electrolysis can remove electrons by direct
chemical reaction.85 And in dye-sensitized solar cells, high
electron lifetimes (2  102 s) generally require iodide to remove
photoholes.79 Augustynski found that TiO2 nanocrystals undergo
self-doping under illumination which increases the conductivity
of the film.97 Surface modifications of TiO2 with phosphate were
also found to increase e/h separation and photoelectrochemical
water oxidation activity.108 This was mainly attributed to the
electrical field eect of the negatively charged anion, which
helped e/h separation. A similar improvement was observed
when SrTiO3 was added to TiO2 nanotubes, a possible result of
formation of a heterojunction.109 Enhanced charge transport
and photocurrents were also observed with TiO2 nanocrystals
on single walled carbon nanotubes as a high surface area
conductive support.110
Closely connected to slow charge transport and recombina-
tion are the slow water oxidation kinetics of TiO2. On the basis
of transient absorption spectroscopy, Durrant reported a milli-
second to second timescale for water oxidation in TiO2.68,69 In
flux dependent measurements, the maximum activity occurred
when four photoholes were present per nanoparticle, in agree-
ment with the stoichiometry of the water oxidation reaction.69
Interestingly, the water oxidation kinetics were independent of
temperature in TiO2 (but not in Fe2O3 photoelectrodes), which
was attributed to the strongly oxidizing valence band edge in
TiO2.111 The kinetics can be improved through the addition of
cocatalysts, as shown in the early examples of TiO2-based
photochemical diodes.112 Here, TiO2 was employed with or
without Ru bipy sensitization,33,113,114 and RuO2 and Pt were
used as cocatalysts for water oxidation and reduction respec-
tively (Fig. 2, above). This configuration was designed to
promote electron injection into platinum, the site for water
reduction, and hole injection into IrO2, where water oxidation
was expected to occur. While hydrogen formation from
sacrificial donors was clearly possible with these systems, the
This journal is c The Royal Society of Chemistry 2013
View Article Online
Chem Soc Rev
initial findings of stoichiometric oxygen evolution in water were
not reproducible.30,32 Initially this problem was attributed to
the back reaction of photoexcited TiO2 with the generated O2,
leading to the formation of superoxide and peroxide species.115,116
But later it was found that the peroxides form by incomplete
oxidation of water,117 and that oxygen formation could not
proceed in the absence of sacrificial electron acceptors. From
todays standpoint, the inability of these systems to photolyze
water must be attributed to the lack of a suitable thermodynamic
driving force. Water electrolysis requires a potential of 1.6 V
(1.23 V plus about 0.4 V in overpotentials), which a nanoscale
catalyst cannot produce without a mechanism for ecient e/h
separation. A similar situation is encountered in nanoscale TiO2
derivatives, such as H2Ti4O9 nanosheets.118 These 1 nm thin
nanocrystals evolve H2 and O2 only in the presence of sacrificial
agents, and oxygen evolution is slow (3.5 mmol h1/100 mg of
catalyst) even in the presence of silver nitrate as electron acceptor.
Electrochemical measurements reveal that the system is limited
by large overpotentials (0.40.85 V at 0.5 mA cm2) for the water
redox reactions, and by a conduction band edge (0.53 V vs. NHE
at pH = 7) that is barely reducing enough for proton reduction. In
very recent work, Maos group reported the formation of black
TiO2 by hydrogenation of B7 nm rutile TiO2 nanocrystals at
200 1C under 20 bar of H2 in a steel autoclave.119 After attachment
of a Pt cocatalyst, 20 mg of this material supported high rates of
H2 evolution (0.2 mmol h1) from a 1 : 1 (vol) methanolwater
mixture exposed to natural sunlight. The findings were attributed
to bandgap narrowing resulting from the introduction of H
defects and disorder into the TiO2 lattice. However, as recent
results on TiO2 nanotubes suggest, it is possible that metal
impurities, in particular Cr from the steel autoclave may play a
role as well.120
ZnO
As a photoelectrode material ZnO is limited by a large bandgap
that prevents visible light absorption, and by chemical instability
under both anodic and cathodic bias. Kyoung Shin Chois group
demonstrated recently that the cobalt oxide-phosphate cocatalyst
can be photodeposited into ZnO rod arrays (Fig. 15) and that this
improves the photocurrent onset potential by 0.23 V.121
Photodeposition was advantageous over electrodeposition,
as it deposits the cocatalyst at locations with high photohole
concentrations. Sun et al. demonstrated improved cathode
photocurrents from n-type ZnO nanowire arrays on p-Si wafers.
The increased photocurrent was attributed to enhanced e/h
separation at the n-ZnO/p-Si interface, and due to reduced light
scattering from the nanorods.122 Previously, Jin Zhangs group
had demonstrated that non-sensitized nanoparticulate ZnO
photoanodes supported photocurrents up to 175 mA cm2
at 1.6 V (NHE) applied potential at pH = 7.4 and AM 1.5
illumination.123
Niobates
While Nb2O5 is not an active photocatalyst for water split-
ting,124 many niobates do split water upon UV irradiation.
One of the best studied materials, K4Nb6O17, owes its high
Chem. Soc. Rev., 2013, 42, 2294--2320 2301

Chem Soc Rev
Fig. 15 ZnO nanorod array after electrodeposition of CoPi catalyst. Reproduced with permission from ref. 121. Copyright 2009, National Academy of Sciences of the
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
United States of America.
activity to its layered structure featuring Nb6O174 anionic slabs
with interlayers that become accessible to water after exchange
of the potassium ions with protons.124 Chemical exfoliation of
this material and the similarly structured Dion Jacobsen phase
KCa2Nb3O10 produces 1 nm thin single-crystalline nanosheets
or scrolls (Fig. 16) that can be dispersed in water. Due to their
homogenous structure and morphology, these nanocrystals and
their Pt and IrO2-modified derivatives are an excellent system for
systematic studies on water photoelectrolysis.125127 For example
it was shown that in pure water, the catalytic ability of the
nanosheets is limited by peroxide formation on the surface. The
peroxides form in 1 : 1 mole ratio with hydrogen.128 This behavior
is reminiscent of TiO2 (see above) and shows that the nano-
crystal-Pt composited are not capable of produce a sucient
photovoltage for complete water photoelectrolysis.
One reason is the very high water oxidation overpotential,
that is not lowered much even after attachment of IrO2 nano-
particles cocatalyst. Another reason is the lack of a charge
separation pathway in the material. According to a photo-
labeling study, electrons and holes are accessible over the
entire nanosheet surface and the edges.130 As a result, charge
carriers in nano-niobates decay on nanosecond timescale.129,131
In the presence of methanol, the nanosheets become very active
photocatalysts for hydrogen evolution under UV light, espe-
cially in the presence of Pt cocatalysts, which reduce the proton
reduction overpotential.108110,125 Under these conditions it is
Fig. 16 Single crystal niobate nanosheet and nanoscroll photocatalysts.126,127,129
2302 Chem. Soc. Rev., 2013, 42, 2294--2320
View Article Online
Review Article
found that the nanoscale sheets are about 16 times as active as
the non-exfoliated parent material. That can be understood in
terms of the reduced transport limitation for electrons and
holes, which in the case of the sheets are generated near the
surface.132 This enhancement persists in HCa2Nb3O10
nanosheets after restacking with KOH or NaOH. The specific
surface area of this composite is ten times greater than in the
condensed phase.133 For restacked nanosheets of HCa2Nb3O10,
HSr2Nb3O10 and HLaNb2O7 it was found, that the activity of the
triple-layered nanosheet (HSr2Nb3O10) is an order of magnitude
higher than that of the double-layered HLaNb2O7, but the
reason for that was not clear.134 A major limitation for solar
energy conversion with HCa2Nb3O10 nanosheets is the 3.5 eV
bandgap that restricts absorbed wavelength to o350 nm. Inter-
estingly, while doping of rhodium into the lattice does move
the absorption edge to about 600 nm, Rh-doped HCa2Nb3O10
nanosheets, gain virtually no activity for visible light driven H2
evolution.135 That demonstrates that charge carriers produced
by excitation of rhodium-based states are not available for the
water redox reaction, even though the Rh-dopants are located
directly at the nanosheet water interface. A more eective way
to achieve visible light driven photocatalysis is by sensitization
of niobate, titanate, titanoniobate nanosheets and nanoscrolls
with Ru bipy dyes, as shown in Fig. 17.
Photogenerated electrons at the Ru bpy complex are quickly
injected into the conduction band and delivered to the Pt
This journal is c The Royal Society of Chemistry 2013

Review Article
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
Fig. 17 Visible light driven H2 evolution from Ru bipy-modified nanosheets and
scrolls in the presence of EDTA. Reproduced with permission from ref. 136.
Copyright 2009, American Chemical Society.
cocatalyst where they reduce water. The holes are quenched by
the sacrificial agents, EDTA, or iodide.89,136139 Restacking of
the nanosheets does not significantly reduce activity, because
the large surface area is preserved.140 The quantum yield for
sensitized H2 production is primarily governed by the electron
injection eciency from the excited-state sensitizer into the
nanosheet. Because of the quasi-molecular properties of the
nanosheets, it is easy integrate them into films on flat and
curved substrates.50,141,142 When electrostatically attached to
SiO2 microspheres, niobate nanosheets continue to catalyze
hydrogen formation from aqueous methanol, with negligible
loss of activity.141 This demonstrates the possibility of sup-
ported photocatalysts for large scale energy conversion. In the
presence of the sensitizers, such supported systems produce H2
with apparent quantum yields of up to 2025% at l >
420 nm.50,142 The dependence of the hydrogen evolution rate
on Pt loading suggests that the scavenging of conduction band
electrons by Pt is substantially faster than charge recombina-
tion or EDTA reduction of the oxidized sensitizer. For shell-type
assemblies of alpha-ZrP or HTiNbO5 sheets, quantum yields of
4761% were achieved.50 Because they integrate light absorbing,
Fig. 18 Types of nanostructured WO3 films. (A) Particles, (B) nanowires, (C) plates. Reproduced with permission from ref. 148 and 149. Copyright 2009, International
Association of Hydrogen Energy, and 2011, Elsevier Science Ltd.
This journal is c The Royal Society of Chemistry 2013
View Article Online
Chem Soc Rev
charge relaying, and redox active components, these assemblies
are termed artificial leaves.
WO3
The use of WO3 for photoelectrochemical water oxidation was
first reported in 1976 by Hodes.143 Compared to TiO2, the
photocurrent density with WO3 is small (usually o3 mA
cm2),144 prompting attempts to increase its surface area by
nanoscaling. Augustynski reported that during photoelectro-
chemical methanol oxidation, nanostructured WO3 electrodes
posses enhanced visible light absorption and current.145,146
The group also observed that after a threshold potential of
about 0.5 V, the current became independent of the applied
potential, in contrast to a bulk-WO3, indicating the absence of a
continuous space charge layer in the nanostructured film.
Photoanodic water oxidation with photocurrents of up to
2.7 mA cm2 under AM 1.5 and 1.0 V applied bias can be
achieved with such electrodes in the presence of methylsulfonic
acid (CH3SO3H) as supporting electrolyte.147 In aqueous H2SO4
and HClO4, the nanoparticle WO3 photoanode degrades over
time and persulfates and hydrogen peroxide are formed as
products. The performance also strongly depended on the
electrolyte concentration. The incorporation of metal ions
(Mo5+, Si4+, Sn4+, Ru3+, Li+) into the WO3 nanocrystals generally
reduces the photocurrent to values below the non-doped film
(2.53 mA cm2, AM 1.5, 1 M H2SO4, EApp = 1.0 V vs. NHE). This was
attributed to the introduction of defects and lower crystallinity of
the film.148
Enhanced photocurrents (26%) for isopropanol oxidation
were also observed with mesoporous WO3, due to the increased
specific surface area.150 Relatively high photocurrents
(B2.0 mA cm2, AM 1.5, 1.0 V vs. NHE, 0.1 M Na2SO4) during
water oxidation were observed for films of hydrothermally
grown WO3 crystal plates of 50200 nm thickness.151 Ohtanis
group compared the performance of flat and vertically aligned
WO3 nanoplates. The vertical ones supported higher photocur-
rents of up to 1.8 mA cm2 (1.2 V vs. NHE, 0.1 M Na2SO4), due to
improved charge transport to the back contact.152
An increase of the photocurrent by up to 60% (to 2.1 mA cm2,
1.5 V vs. NHE, 1 M H2SO4, AM 1.5) was observed after plasmon
enhancement with silver nanoparticles (see below).153 Sunkara
and coworkers prepared WO3 nanowire electrodes by thermal
deposition onto FTO electrodes (Fig. 18).149 These gave water
oxidation photocurrents of B1 mA cm2 at 1.0 V vs. NHE in 0.5 M
Chem. Soc. Rev., 2013, 42, 2294--2320 2303

Chem Soc Rev
H2SO4 (AM 1.5). Tungsten nitride nanowires (WN2) could be
prepared similarly by treatment of WO3 nanowires with NH3.149
The photocurrents from these were slightly higher, but the
material undergoes anodic photocorrosion.
The performance of WO3 is also limited by the slow kinetics
of water oxidation, which occurs on the millisecond to second
timescale.154 Cocatalysts are expected to lessen this problem by
reducing the kinetic barrier. Interestingly, while photodeposited
Co-phosphate oxide (Co-Pi) on WO3 improves the photocurrent,
this does not appear to be a result of improved redox reactions,
but due to improved e/h separation.155 Also, it was found that the
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
Co-Pi layer reduces photocorrosion of WO3. Suspended powders
of WO3 can photocatalyze water oxidation in the presence of a
chemical bias.156 This reactivity is preserved in single crystalline
0.75 nm thin nanosheets of WO3 obtained by chemical exfoliation
of the layered Bi2W2O7. The nanosheets have an increased band
gap from quantum confinement and they are prone to photo-
corrosion in neutral solution, similar to the bulk material.157
Cu2O
As a p-type semiconductor with a bandgap of 2.0 eV, Cu2O
functions as photocathode for water reduction. Only a few
applications of nanoscale Cu2O to water photoelectrochemistry
are known today, because the material undergoes oxidative
decomposition in air and anodic or cathodic photocorrosion
(formation of CuO or Cu). Gra tzel and coworkers recently
demonstrated that a combination of 4 nm Al2O3 and of a
11 nm TiO2 overlayer can reduce photocorrosion (Fig. 19). After
deposition of Pt nanoparticles, water reduction proceeded at
EApp = 0 V vs. NHE with photocurrents of up to 7.6 mA cm2
(AM 1.5) in 1.0 M Na2SO4 solution.158
By controlling electrochemical deposition parameters, Choi
and coworkers synthesized dendritic Cu2O films that produce
anodic currents of B0.25 mA cm2 in 0.2 M K2SO4 (100 mW, l >
400 nm), that are likely aected by photocorrosion, and not
necessarily due to water oxidation.159 Turner and coworkers
synthesized CuO/Cu2O nanowires by electrochemical deposition.
Even though the composition be well controlled, the nanowires
only supported small photocurrents (10 mA cm2).160
BiVO4
As a ternary compound, BiVO4 had only recently been discovered
as a photoanode material for water oxidation.161 Due to the
positive potential of the conduction band edge it requires an
electrical or chemical bias. A microporous Mo-doped BiVO4
electrode was prepared using a surfactant assisted metalorganic
Fig. 19 Surface-protected Cu2O photocathode. Reproduced with permission
from ref. 158. Copyright 2011, Nature Publishing Group.
2304 Chem. Soc. Rev., 2013, 42, 2294--2320
View Article Online
Review Article
decomposition technique at 500 1C.162 A stable photocurrrent
density of 1.0 mA cm2 at 1.0 V vs. Ag/AgCl was achieved under
standard AM 1.5 illumination using 0.5 M Na2SO4 aqueous
solution in phosphate buer at pH 7. Nearly double photo-
current density was observed at 1.0 V vs. Ag/AgCl for a Co-Pi/
BiV0.98Mo0.02O4 composite photoelectrode.163 Similar enhance-
ment due to the presence of Co-Pi was observed in mesostructured
SiO2BiVO4 composites.163 Here SiO2 increases the optical trans-
parency of the electrode and possibly improves hydroxyl adsorp-
tion. Photodeposited Co-Pi gave higher photocurrents at less
positive applied potentials than electrodeposited Co-Pi, likely due
to better proximity to the BiVO4 photoactive sites. Mullins group
prepared nanostructured BiVO4 films by co-evaporation (reactive
ballistic deposition) of bismuth and vanadium in an oxygen
atmosphere.103 Films deposited with excess vanadium (V/Bi = 2)
showed the highest initial photocurrents with IPCE values above
21% for light wavelengths of 340460 nm (in 0.5 M Na2SO4 at 1.0 V
vs. Ag/AgCl), but leached vanadium into the electrolyte. Stoichio-
metric films were stable, but their photocurrents were limited by
surface reaction kinetics. The addition of cobalt as an electrocata-
lyst increased the photocurrent three times. In very recent work,
Choi and coworkers improved the water oxidation kinetics of
BiVO4 by photodepositing a nanostructured 100 nm thick layer
of FeOOH. This optimum layer thickness was identified by
monitoring the water oxidation photocurrent during deposi-
tion.164 To date, the largest photocurrent density with BiVO4 of
1.7 mA cm2 at 1.23 V applied potential (vs. RHE), was observed
by Roel van de Krols group for a BiVO4 film modified with 30 nm
of Co-Pi.165 The performance was attributed to good charge
separation and low recombination rates. At high illumination
intensity, the eciency was limited by transport of the majority
carriers.
dx systems
Metal oxides with partially filled d-shells typically exhibit low e/
h lifetimes. In addition, the first row transition metal oxides
also suer from low e/h mobility they are classified as Mott
insulators.52 In theory, both problems can be addressed
through nanoscaling, and so dx systems have been intensively
studied as photoelectrodes. The best known metal oxide here is
Fe2O3, but reports on improved photocatalytic properties with
nano-MnO2 and nano-IrO2 have appeared as well.
Fe2O3
The early results on photoelectrochemical water oxidation with
Fe2O3 date back to experiments by Allen Bard in 1976,166
Karl Frese in 1978,167 and Andrew Hamnett in 1983.47 This work
established the electrochemical stability of Fe2O3 under anodic
conditions, but also showed that the material produced low
photocurrents. It is now clear that these are due to a combi-
nation of short exciton lifetime (B10 ps),4,5 poor charge carrier
mobility,6 medium light penetration depth (a1 = 118 nm at
l = 550 nm),7 and slow water oxidation kinetics. For higher
performance, nanostructuring is employed as a means to
increase the surface area, doping to improve charge transport,
and surface modifications to improve e/h separation and water
This journal is c The Royal Society of Chemistry 2013

Review Article
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
Fig. 20 Nanostructured Fe2O3 films made from ultrasonic spray pyrolysis (left),174 by air-sintering of Fe3O4 colloids (middle),183 or by calcination of beta-FeOOH
colloidal rods.179 Reproduced with permission from ref. 174, 179 and 183. Copyright 2006, 2010, American Chemical Society. Copyright 2005, John Wiley and Sons.
oxidation kinetics. This multidimensional strategy has been
proven successful for Fe2O3 and can serve as a template for the
optimization of other materials.
Syntheses
In 1984 Bockris and coworkers introduced spray pyrolysis
for the synthesis of nanostructured Fe2O3 electrodes.168 This
technique was further developed by Shahed Khans group169,170
and by Jan Augustynski.171,172 Fe2O3 gained popularity after
Michael Gra tzels group modified the spray pyrolysis method in
2005 to obtain Fe2O3 films with fractal morphologies,173175
partially due to incorporation of silicon impurities.176 Other
nano-morphologies were obtained by surface deposition and
oxidation of preformed Fe3O4 nanocrystals,177 by depositing
Fe2O3 inside of a SiO2 porous template,178 and by coating FTO
electrodes with beta-FeOOH colloidal rods followed by calcina-
tion (Fig. 20).179 Active Fe2O3 nanowire arrays can be obtained
by solution growth on ITO,180 or plasma oxidation of iron
metal, followed by removal of an amorphous FeO overlayer
with HF.181 Columnar nanoscale structures are accessible by
reactive ballistic deposition in which iron metal is evaporatively
deposited on a cold substrate in an oxygen atmosphere.182
Among these morphologies, the fractal Fe2O3 shape can achieve
photocurrents of up to 2.2 mA cm2 in 1 M NaOH (pH = 13.6)
under AM 1.5 illumination and 1.23 V applied bias versus
relative hydrogen electrode (+0.43 V on the NHE scale).
For sintered Fe2O3 spheres, the photocurrent under similar
conditions is 0.6 mA cm2,183 but only after annealing at 800 1C,
Fig. 21 Fe2O3 nanowire arrays from plasma oxidation of iron181 and nanostructured Fe2O3 on TiS2 nano-net.184 Reproduced with permission from ref. 181 and 184.
Copyright 2012, IOP Science. Copyright 2011, American Chemical Society.
This journal is c The Royal Society of Chemistry 2013
View Article Online
Chem Soc Rev
which is accompanied by Sn4+ doping of the Fe2O3 film. Nano-
rods have given up to 0.1 mA cm2 at 1.23 V versus RHE in 1 M
KOH,181 and with nano-Fe2O3 on high surface TiS2 substrates,
Wang and coworkers were able to achieve 1.6 mA cm2 at 1.23 V
RHE and AM 1.5 in 1 M NaOH (Fig. 21).184
The enhanced performance of fractal-Fe2O3 is due to a
combination of large surface area for eective hole transport
and large film thickness for improved light absorption.174 The
performance can be improved further through a combination
of dierent methods. One is to improve electronhole separa-
tion in the material and slow down e/h recombination. Another
one is to improve the water oxidation kinetics with cocatalysts.
A third one uses dopants to improve electron transport in Fe2O3
to the back electrode. And a last method seeks to modify the
Fe2O3 energetics for enhanced charge transfer. Plasmonic
nanostructures to enhance light absorption have not been
eective so far (see section at end).
Electronhole separation
At large applied potentials, the performance of Fe2O3 is no
longer limited by the kinetics of water oxidation but more by
bulk recombination.185 Therefore, attempts were made to
increase electronhole separation in the material. Based on
intensity-modulated photocurrent spectra, Laurence Peters
group concluded that the improved performance of Cobalt(II)
nitrate treated Fe2O3 photoanodes was not due to enhanced
kinetics, but due to improved e/h separation at the surface and
reduced e/h recombination.186 Choi and coworkers treated
Chem. Soc. Rev., 2013, 42, 2294--2320 2305

Chem Soc Rev
Fe2O3 with Zn and Al nitrates to generate thin ZnFe2O4,
ZnFe2xAlxO4 or Fe2xAlxO3 overlayers. On the basis of Mott
Schottky plots, the improved performance was not due to
changes in kinetics or energetics, which supported a hypothesis
of improved electronhole separation.187,188 Kevin Sivulas
group investigated the eect of a 2 nm Nb2O5 underlayer on
the performance of Fe2O3 films. The layer seemed to improve
e/h separation at the Fe2O3FTO interface, thereby raising the
photocurrent.189 When a 0.12.0 nm thin Al2O3 overlayer was
coated on top of the Fe2O3, the photocurrent onset shifted to
less positive potentials, which was attributed to the passivation
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
of hole trap sites on the Fe2O3 surface.190 A similar eect was found
for Ga2O3 overlayers.191 Based on electrochemical impedance
studies, Cesar et al. discussed the possibility that Si dopants in
the Fe2O3 lattice increased the carrier density from 1017 cm3 to
1020 cm3. Theoretically, the concominant reduction of the
space charge layer width could allow field-assisted electron/
hole separation in even small Fe2O3 grains. Alternatively, the
observed enhancement could have been due to improved hole
transfer related to the increase in surface area.76 While a thin
layer of the Co-Pi catalyst on Fe2O3 improves the water oxida-
tion kinetics (see below) it was also found to increase the hole
lifetime, indicating that it helped in separating e/h pairs.192
Water oxidation kinetics
It is well established that slow water oxidation kinetics limit the
performance of Fe2O3,47,166 especially at low applied potentials.185
This was recently confirmed with electrochemical impedance
measurements193,194 in Thomas Hamanns group, by comparing
electrode kinetics of Fe2O3 in water and with iron hexacyanide
complexes.195 Recent transient absorption measurements show
that water oxidation on Fe2O3 occurs on the 103 to 1 second
timescale.196,197 In contrast to TiO2, the water oxidation rate on
Fe2O3 is also found to depend on temperature.111 The overpotential
for water oxidation can be lowered by adding cocatalysts. IrO2 is
among the most active water oxidation electrocatalysts, and indeed,
electrophoretic modification of Fe2O3 with 2 nm IrO2 nanoparticles
allowed photocurrents of 3.0 mA cm2 at an applied potential of
1.23 V (vs. RHE).46 This is the highest performance of any Fe2O3
photoelectrode today. A smaller increase of the photocurrent was
also observed by Gamelin and coworkers, after electrodeposition of
the Co-Pi catalyst onto Fe2O3 (Fig. 22). In this case, the cocatalyst
also shifted the photoonset potential to less positive potentials by
about 0.3 V.198 When the Co-Pi deposition was photoassisted, a
more uniform coating was obtained, and photocurrents compar-
able to that of IrO2 (2.8 mA cm2 at 1.23 V vs. RHE) were
possible.199 Remarkably, simple dipping of the Fe2O3 into cobalt
nitrate solution also led to substantial increases of the photocurrent
and cathodic shifts of the current onset potential. The increasing
eectiveness in going from Co2+ dipping to Co-Pi electro- and
photodeposition was confirmed in independent work in Kyoung
Chin Chois group.200 Surface adsorption of Co2+ ions was also
eective in improving the performance of Zn and Al modified
Fe2O3 photoanodes.187,190
While it is tempting to attribute the benefit of Co2+ surface
treatment to improved water oxidation kinetics, intensity
2306 Chem. Soc. Rev., 2013, 42, 2294--2320
View Article Online
Review Article
Fig. 22 Fe2O3 film before and after photodeposition of Co-Pi catalyst. Repro-
duced with permission from ref. 198. Copyright 2009, American Chemical
Society.
modulated photocurrent spectroscopy (IMPS) by Peters group
show that Co2+ surface ions suppress e/h surface recombina-
tion.194 However, a similar lowering of the e/h recombination
rate by three orders of magnitude, presumable by aiding
e/h separation, was also observed for the established water
oxidation catalyst Co-Pi on Fe2O3, based on transient absorp-
tion spectroscopy data.192 Lastly, work on porous Fe2O3 reveal
that residual SiO2 from a porous structure template can retard
the water oxidation kinetics, and move the onset potential to
more positive values.178
Charge transport
As a Mott insulator, Fe2O3 suers from to slow electron and
especially hole transport.193,201 Hole transport to the surface is
usually improved through nanoscaling the material, but elec-
trons still need to travel to the back-contacts of the Fe2O3 film.
Thus, research has focused on improving electron transport by
using dopants. Work in Augustynskis group showed that
dopants, incl. Ni2+, Al3+, Ti4+, Ta5+, In3+, Pt4+ have a strong
influence on the photoanode performance of Fe2O3.172 Most of
these ions reduce the photocurrent, but a beneficial eect was
noted for either Al3+ (1%) or Zn2+ (4%) in combination with Ti4+
(5%). These modifications lead to a favorable shift of the
photocurrent onset potential, which the authors attributed to
an increase of the hole diusion length. An additional correla-
tion exists between the crystallinity of the films, monitored by
Raman spectroscopy, and their photoactivity. The photoactivity
of iron oxide was improved by co-deposition with Mo or Cr.
Because there were no electrocatalytic eects of these ions,
the observed increase of photocurrent was attributed to
better charge transport.202 Similar observations were made with
This journal is c The Royal Society of Chemistry 2013

Review Article
Pt4+-doped Fe2O3203 and with Sn4+-doped Fe2O3.183 Here, the
Sn4+ ions either diused into the Fe2O3 film the FTO substrate
during annealing at 800 1C, or were intentionally added as
SnCl4. Grimes and Mallouk used MottSchottky technique to
quantify carrier concentrations in Sn-doped Fe2O3 nanowire
arrays (800 and 2500 nm long). Increasing temperature lead to
increased concentrations from 7  1019 to 2.45  10 20 cm3
and higher photocurrents of up to 0.7 mA cm2 at 0.43 V vs.
NHE (=1.23 RHE) in 1.0 M NaOH and AM 1.5.204 Yat Lis group
used a similar analysis to measure 1.89  1019 cm3 for the
electron concentration in Sn-doped Fe2O3 films obtained by
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
treatment with SnCl4 and annealing at 650 1C.205 Turner and
coworkers prepared Ti-doped Fe2O3 films by co-sputtering of
the corresponding oxides. Based on Mott Schottky plots, the
carrier densities (1.81  10142.43  1014 cm3) increased with
Ti content, but surprisingly, were lower than in undoped Fe2O3.
The dopants also raised the photocurrent, which was attributed
to improved carrier transport and a reduced bandgap of Ti :
Fe2O3.206 The same group found a similar eect for MgTi
co-doped Fe2O3. Here, the optimum performance was seen at
1 : 1 dopant ratio, which was attributed to reduced charge
defects, due to the compensating influence of the +2 and +4
dopant charges.207 Titanium-doped Fe2O3 films prepared from
tetrabutoxy-titanium and iron chloride, showed higher photo-
currents, due to enhanced charge transport resulting from
higher majority carrier concentrations (1.36  1019 cm3),
one hundred times above that of non-doped Fe2O3.208 A similar
eect and photocurrents of up to 2.4 mA cm2 (1.0 M NaOH at
1.23 V NHE, 1.5 AM illumination) were observed upon doping
Cd(2+) ions into the Fe2O3 films.209 An improvement of the
visible light response and the water oxidation kinetics were
also noted.
Energetics and density of states
The energetics of a semiconductor are central to its function as
a photoelectrode. For Ti/F co-doped Fe2O3 films made by
electrochemical codeposition of FeCl3 and Ti(OiPr)4 in the
presence of KF, followed by annealing, McFarlands group
observed a 0.2 V photoonset shift to 0.6 V (NHE, in 1 M
NaOH), after surface treatment with CoF3.210 This was attrib-
uted to a shift of the Fe2O3 flatband potential due to fluoride
incorporation as TiF into the film. No change of the flatband
potential was observed in the absence of Ti. Using Resonant
Inelastic X-ray scattering (RIXS), Vayssieres recently observed a
widening of the egt2g gap of Fe3+ in films of Fe2O3 nanorods
(10 nm diameter). This was interpreted as a result of a one-
dimensional quantum size eect.179
Suspended nanostructures
The majority of work on nano-Fe2O3 involves thin films, but a
few papers on water photoelectrolysis with suspended nano-
tzels
particles have also appeared in the literature. In 1987, Gra
group reported the formation of hydrogen from irradiated
colloidal Fe2O3 in pure water.211 Similar observations were
made by Gondals group, under Laser excitation.212 As the
conduction band of Fe2O3 is not negative enough to allow
This journal is c The Royal Society of Chemistry 2013
View Article Online
Chem Soc Rev
Fig. 23 Left: silver nanoparticles formed on nano-Fe2O3 after irradiation in
aqueous AgNO3. Right: correlation between Fe2O3 particle size and O2 evolution
rate. Reproduced with permission from ref. 213. Copyright 2011, Royal Society of
Chemistry.
proton reduction, the formed hydrogen was likely due to radical
processes following the photoreduction of O2, and involving
carbon impurities in these system.211 In 2012, our laboratory
reported oxygen formation from Fe2O3 nanocrystal dispersions
in water and 0.1 M AgNO3 as sacrificial electron acceptor. Even
though the performance of the system was limited by silver
deposition on the Fe2O3 nanocrystals (Fig. 23) a systematic
increase of O2 production rates could be seen with decreasing
particle size. This was attributed to improved hole extraction in
the smaller nanocrystals.213
IrO2
Because IrO2 has a very low electrochemical overpotential for
water oxidation,214219 it is widely used in commercial water
electrolyzers and as a cocatalyst in combination with a photo-
electrode (see below). In 2011, Osterloh and coworkers
described visible and UV-light induced photocatalytic oxygen
evolution from IrO2 sols in the presence of silver nitrate or
sodium persulfate as sacrificial acceptors.220 The photocatalytic
activity is unusual, considering the short lifetime of photo-
generated holes in metallic materials, incl. IrO2. In the present
case, photohole extraction was possible due to the small size of
the particles (2 nm).
MnO2
In bulk form, MnO2 is not active for photoelectrochemical water
oxidation,218 although it does function as electrocatalyst.221,222
Recently, several reports suggest that nanostructuring activates
birnessite-MnO2 for photocatalytic water oxidation. In 2005,
Sakai and coworkers reported small photocurrents for a film of
0.5 nm thin nanosheets of MnO2 on FTO in acetonitrile with
LiClO4 as electrolyte and a Pt counter electrode. The origin of
the current was not clear, although the nanosheets remained
stable.223 Small photocurrents were also observed for this
material on gold, in the presence of methanol.224 Later, Leone
Spiccias group reported photoelectrochemical water oxidation
(3 mA cm2 for electrodes at 1.20 V vs. NHE in 0.1 M Na2SO4) on
Nafion/glassy carbon membranes with embedded manganese
oxide clusters.225 These clusters were later found to convert
into birnessite-MnO2 nanocrystals (Fig. 24).226 The conversion
of tri-azacyclononane- and terpyrididyl-stabilized Mn4O5 and
Mn4O6 clusters into MnO2 nanoparticles was demonstrated
Chem. Soc. Rev., 2013, 42, 2294--2320 2307
 View Article Online

Chem Soc Rev Review Article

Metal chalcogenides
Because they show quantum confinement eects, metal
chalcogenides, especially those of the group 12 metals, have
received much attention in the nanoscience community.
However, photocorrosion makes these materials unsuitable
for water oxidation, and so metal chalcogenide nanocrystals
are mainly used for visible light sensitization of metal oxide
photoelectrochemical cells,232 and for basic science studies on
photocatalytic water reduction, interfacial charge transport
and233 and quantum confinement.62,234 Countless papers have
focused on CdS nanoparticles and nanocomposites as visible
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.

light driven photocatalysts for proton reduction in the presence

of sacrificial agents, including EDTA, sodium hydrosulfide, and
Fig. 24 Nafion film showing presence of manganese-oxide nanoparticles
formed after electro-oxidation of [Mn4O4L6]. Reproduced with permission from
methanol.235 However, because the intrinsic photocatalytic
ref. 226. Copyright 2011, Nature Publishing Group. proton reduction activity of bulk CdS is already very high
(quantum eciencies of 2534%),236238 nanostructuring is
not eective for increasing performance. The situation is
independently in Gary Brudvigs group.227 When embedded
similar for bulk ZnS, where a H2 evolution quantum eciency
inside of a Nafion membrane, or coated onto FTO, these
of 90% was measured by Reber, Meyer and Rusek in 1984.236
nanoparticles supported visible light (500 mW cm2) driven
Among group 12 chalcogenides, cadmium selenide, on the
water oxidation with photocurrents of up to 16 mA cm2 at
other hand, is interesting because its bulk form does not
applied potentials of 1.2 V vs. NHE. Thomas Jaramillos group
catalyze hydrogen evolution from water without an applied
showed that Na-birnessite-type MnO2 thin films made via elec-
bias, due to a high proton reduction overpotential and a non-
trodeposition could support photoelectrochemical water oxida-
favorable conduction band edge.239 The energetics of this
tion at large applied potentials (>1.45 V vs. RHE, 0.1 M NaOH)
material can be improved significantly through nanoscaling,
but the photocurrents were very small (o50 mA cm2).228
taking advantage of the quantum size eect. Nanoribbons of
VO2 CdSe become active proton reduction photocatalysts in the
presence of Na2SO3/Na2S as sacrificial agents.240,241 For CdSe
Nano-VO2 has been studied extensively due to its metal insu- quantum dots, Holmes et al. recently described an exponential
lator transition at 67 1C, and as a potential battery material. dependence of the hydrogen evolution quantum eciency
There is one report on photoelectrochemical water reduction (from sodium sulfite solution) on the band gap of the dots
with VO2 nanorod arrays on silicon, under UV irradiation. As (Fig. 25).62 This dependence is expected from MarcusGerischer
the conduction band edge of VO2 would be expected to be to theory, which correlates the proton reduction rate with the
low for proton reduction, it is likely that H2 evolution in this thermodynamic driving force, given by the dierence between
system is not from VO2 but from photocorrosion of the silicon the CdSe flatband potential and the proton reduction potential.
wafer.229 Size and shape dependent electron injection from CdSe dots into
Metal nitrides metal oxides has also been observed by Prashant Kamats group,
using transient absorption spectroscopy and emission quenching
Because the electronegativity of nitrogen is lower than that of experiments.61,234,242,243 By changing the CdSe particle diameter
oxygen, metal nitrides are prone to anodic photocorrosion.
However, they have considerable potential as photocathodes.
Indeed, W2N nanowire arrays formed by nitridation of WO3 are
found to support anodic photocurrents of 0.2 mA cm2, but
performance rapidly decays due to formation of the oxides.
Improved stability was found for the mixed phase W2NWO3.149
Recently, Yuliati et al. reported the synthesis of Ta3N5 nano-
crystals by nitridation of Ta2O5 using C3N4 as a structure
template. After attachment of Pt cocatalyst particles, the
Ta3N5 was active for photocatalytic hydrogen formation from
aqueous methanol under visible light irradiation. The activity
was higher than that of bulk Ta3N5.230 In another study, Ta3N5
nanowire arrays were prepared by ammonolysis of Ta2O5 nano-
wire arrays at 800 1C.231 At 450 nm and an applied bias of 0.5 V
these electrodes supported IPCE values of 5.3% in 1 M KOH
solution. The performance improved by a factor of B4 after Fig. 25 Quantum size eect enhanced water reduction. Reproduced with
addition of IrO2, Co3O4 or Co-Pi nanoparticle cocatalysts. permission from ref. 62. Copyright 2012, Royal Society of Chemistry.

2308 Chem. Soc. Rev., 2013, 42, 2294--2320 This journal is c The Royal Society of Chemistry 2013

Review Article
from 7.5 to 2.4 nm, the electron transfer rate increases by three
orders of magnitude.61
Koenenkamp used photoconductance measurements
to analyze quantum size modulated electron injection from
nano PbS into TiO2. No injection occurred when the particle
diameter was above the critical value of 2.5 nm, due to unfavor-
able energetics.244 Parkinsons group recently measured increased
quantum eciencies for PbSe nanocrystal-sensitized TiO2 electro-
des. These were attributed to multiple exciton generation (MEG) in
the quantum dots.70 For nano-Bi2S3 on TiO2, Peter and coworkers
shows that injection of photogenerated electrons into the metal
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
oxide was greatly enhanced in the presence of hydrosulfide.66 This
was attributed to a lifting of the conduction band edge of Bi2S3 by
the potential determining eect of this ion. Homanns group
observed photocatalytic H2 evolution from CdS nanocrystal-
sensitized Ni/NiO-loaded KNbO3 crystals of up to 150 mmol h1
g1 under visible light from aqueous isopropanol. In contrast, the
free dots only had very low H2 evolution activity (5 mmol h1 g1).
The higher rate of the composite was attributed to the presence of
the NiO/Ni cocatalyst.245 Using electrostatic force microscopy, Alivi-
satos group measured the built-in potential of nano-Cu2SCdS
heterostructures. This potential forms when the electrons in the
two phases reach electrochemical equilibrium it is equivalent to
the barrier height of the space charge layer in semiconductor
electrolyte junctions.246
Recently, Kanatzidis and coworkers reported photocatalytic
hydrogen evolution from a Sn2S4Fe4S4 gel after sensitization
with a ruthenium tris bipyridyl complex.247 Even though the
hydrogen evolution rate was very low, it is unprecedented for
this class of compounds.
Main group elements and compounds
Due to its suitable bandgap and high conduction band edge,
silicon is considered an attractive photocathode material for
water reduction. Microstructured silicon films are pursued in
an attempt to replace the ultrapure photovoltaic silicon with
less expensive material. To compensate for the lower minority
carrier diusion length of reagent grade silicon, the electrode
surface area needs to be increased accordingly. This also allows
for the possibility of less expensive water reduction cocatalyst
with lower per atom activity. Nate Lewis Joint Center for
Artificial Photosynthesis fabricated silicon microwire arrays as
photocathodes for H2 production.248,249 The 4060 mm long
microrods consisted of p-type Si grown by vapor liquid solid
process (SiCl4 and BCl3 as dopant) on p-Si wavers. A copper
growth catalyst is preferred over gold, because the latter con-
taminates the silicon with impurities of 1.7  1016 cm3
causing enhanced recombination and lower photovoltage.250
To increase e/h separation at the siliconelectrolyte interface, a
n+p junction was added, and Pt nanoparticles were chemically
deposited at the Si tips to reduce the kinetic barrier for proton
reduction. In 0.5 M H2SO4 and at an applied potential of +0.3 V
vs. NHE such photocathodes reach 15 mA cm2 (AM 1.5), five
times the value of the best Fe2O3 photoanodes. However, the
photocurrent is only half of that of planar n+p junction Si
electrodes (27.5 mA cm2), due increased carrier recombination
This journal is c The Royal Society of Chemistry 2013
View Article Online
Chem Soc Rev
at the higher microrod surface area. While the high aspect ratio
of the microrod photoelectrode enhances minority carrier trans-
port to the semiconductorliquid interface, optimum majority
carrier extraction through the back contact requires precise
control of Si doping.251 Also, it was found that the spacing of
the rods and their 2D arrangement is critical for reducing light
losses that arise from scattering and reflection. While ordered
arrangements did allow higher rod packing density and better
optical absorption, they also made the absorption profiles
anisotropic, leading to dead spots in a device (Fig. 26).252
Another advantage of the microwire arrays is that, after
coating with a polymer, it can be peeled o the silicon waver
and used as a flexible photoelectrode module.253 This poten-
tially allows the construction of milliscale Tandem devices.
Systematic studies on the benefit of adding nanoparticle
cocatalysts are described in the separate section below.254
Peidong Yangs group demonstrated that the cathodic photo-
current of Si microrod arrays is enhanced 5 times after exten-
sion with InGaN nanowire branches.255 Solution-grown
GaP nanowires were also active for photocatalytic hydrogen
evolution (Fig. 27).256
After deposition onto FTO electrodes, the illuminated nano-
wire films gave a photovoltage of up to 150 mV, and were able to
photoreduce water in the presence of methanol as sacrificial
electron donor. Nanowire membranes could be fabricated by
deposition of the nanowires onto poly(vinylidene fluoride)
filters.
Carbon nitride
Even though the compound C3N4 has been known as melon
for over hundred years, its photocatalytic properties were only
recently discovered in Markus Antoniettis group.257 The struc-
ture of the material consists of graphite-like layers in which
some of the carbon atoms are replaced by nitrogen atoms
(Fig. 28). The bandgap and photocatalytic water splitting prop-
erties can be adjusted by copolymerization with barbituric acid,
aminoborane, sulfur, or other heteroatom sources.258260 Based
on surface photovoltage measurements,261 and photoelectro-
chemistry,260 g-C3N4 is an n-type semiconductor that functions
as a weak photoanode material. The low photocurrents suggest
fast e/h recombination and slow charge transport. Therefore,
nanostructuring can be useful for increasing the specific
surface of the compound and its activity.258,262 For example,
while under visible light, platinum-modified bulk g-C3N4
produces hydrogen from aqueous ethanolamine solution at
rates of only 8 mmol h1 g1, mesostructured g-C3N4 reaches
rates of up to 149 mmol h1 g1.262 After integration of Co3O4
nanoparticle cocatalysts, photocatalytic water oxidation with
1.1% QE at 420 nm can be achieved using silver nitrate as
sacrificial electron acceptor.263 While these results are encoura-
ging, prolonged water oxidation is expected to cause significant
photocorrosion. Lastly, microporous C3N4 was also used as a
structure template/nitriding agent for the synthesis of Ta3N5
nanocrystals from TaCl5 and ammonia.230 The resulting nano-
crystals exhibited 10 times higher activity for photocatalytic
proton reduction than Ta3N5 (under visible light and after
Chem. Soc. Rev., 2013, 42, 2294--2320 2309

Chem Soc Rev
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
Fig. 26 Fabrication, morphology, and light scattering properties of Si radial microwire array photocathodes. Reproduced with permission from ref. 248 and 252.
Copyright 2011, American Chemical Society. Copyright 2010, Nature Publishing Group.
Fig. 27 Solution-grown GaP nanowires before (B) and after (D) removal of Ga
cocatalyst. Reproduced with permission from ref. 256. Copyright 2011, American
Chemical Society.
loading of Pt nanoparticle cocatalyst). This was attributed to a
lower defect concentration in the template nanocrystals.
2310 Chem. Soc. Rev., 2013, 42, 2294--2320
View Article Online
Review Article
Nanoparticles as cocatalysts
The systematic study of metal nanoparticles and their potential
as metal microelectrodes goes back to work by Henglein in the
70s.264 Platinum and IrO2 have particularly high activity for
electrocatalytic water oxidation and reduction,265,266 and thus
they were initially employed by Gra tzel and Lehn as water
splitting cocatalysts.267270 In recent years attention has shifted
towards more abundant and less expensive elements that are
more suitable for large scale operation. In addition to lowering
the activation barrier for water redox reactions, the role of
cocatalysts is to assist in e/h separation at the cocatalyst
semiconductor interfaces. Electron injection from excited
TiO2 into gold and silver nanoparticles271,272 and hole injection
into IrO2 can be observed directly with transient absorption
spectroscopy.273 Metalsemiconductor interfaces can form
Schottky junctions with rectifying properties, but at the nano-
scale the electrical fields resulting from such junctions are
greatly diminished. A problem of cocatalystsemiconductor
interfaces is that they generate recombination sites for charge
carries, thereby deteriorating photoelectrochemical perfor-
mance. However, this negative eect is generally overcompen-
sated by improved water redox kinetics.
Pt/Pd/Rh/Ru
Some of the first nanoscale photocatalytic systems employed a
combination of RuO2 colloids for water oxidation, and of Pt or
This journal is c The Royal Society of Chemistry 2013
 View Article Online

Review Article Chem Soc Rev

Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.

Fig. 28 TEM images of porous (A) and bulk C3N4 (B). A photograph and the schematic structure of the tri(s)triazine units are also shown. Reproduced with permission
from ref. 262. Copyright 2009, American Chemical Society.

Ru colloids for water reduction. Ruthenium dyes served as

optical sensitizers, and the methylviologen +/2+ redox couple
as an electron relay between Pt and the dye.274,275 Either oxygen
or hydrogen formation could observed in the presence of
Ce(SO4)2 or and EDTA as sacrificial agents.276
Platinum nanoparticles are most common cocatalysts for
water reduction.39 In 1981, Mills showed that Pt nanoparticles
could be photodeposited onto TiO2 during band gap excitation
in the presence of a solution of chloroplatinic acid and Fig. 29 HR-TEM images of Rh-loaded GaN:ZnO before and after photodeposi-
formaldehyde. Here, the photogenerated electrons reduce the tion of the Cr2O3 shell. Reproduced with permission from ref. 286. Copyright
metal salt, favoring deposition at the electron-rich sites.277 2006, John Wiley and Sons.

Alternatively, Pt nanoparticles could be added by salting out

citrate stabilized Pt nanoparticles in the presence of the semi-
conductor powder. Photocatalysts prepared by the latter proce-
dure were more active for hydrogen evolution in the presence of
aqueous EDTA solution, but the higher activity could have been
due to the presence of residual citrate. More recent studies by
Stefan Bernhards group have demonstrated photocatalytic
water reduction with ruthenium or iridium dye-sensitized
platinum or palladium colloids in the presence of sacrificial
agents.278280 Platinum cocatalysts are sub-optimal for use in
photocatalysts because Pt also catalyzes water formation from
H2 and O2 (the water splitting backreaction). At high H2/O2
concentrations, the backreaction rate becomes equal to the rate
of the forward reaction, and net gas formation ceases. While Au
nanoparticles do not catalyze water formation from H2 and O2,
they can photoreduce O2, and their activity for proton reduction
is about 10 times lower than that of Pt.281 A better way to stop
the back reaction consists in adding a 2 nm Cr2O3 coating onto
the proton reduction cocatalyst, as shown by Kazunari Domens
group (Fig. 29).282,283 The coating is applied to either Rh, Ir, or
Pt cocatalysts, which accept photogenerated electrons from the
semiconductor (in this case GaN:ZnO). Electrons migrate
through the Cr2O3 layer and catalyze proton reduction, but
not oxygen reduction at its surface. The Cr2O3 shell also acts as
oxygen barrier and prevents direct reduction of oxygen at the
metallic nanoparticle core.
An interesting new strategy to obtain active water oxidation
catalysts is to embed highly active metal oxides into soluble
polyoxometallates. This was recently demonstrated by Sartorel
et al.284 and separately by Craig Hills group.285 The decaanion
[Ru4(m-O)4(m-OH)2(H2O)4(g-SiW10O36)2]10 contains a Ru4 cluster
core sandwiched between two [g-SiW10O36]8 units. In the
presence of Ce4+ it generates oxygen with turnover frequencies
of up to 450 s1.284
The related complex [Co4(H2O)2(PW9O34)2]10 is comprised
of a Co4O4 unit sandwiched between two (PW9O34)9 halfs.284
In the presence of the [Ru(bpy)3]3+ as a sacrificial acceptor,
oxygen can be generated at turnover frequencies up to 0.9 s1.
IrO2
In 1991, Andrew Mills published an activity ranking of micro-
crystalline particles for catalytic water oxidation using
cerium(IV) ammonium nitrate as oxidant. The study identified
Ru, RuO2 and IrO2 as most active catalysts.218 The activity of Pt,
Ru, or Os doped IrO2xH2O colloids is even higher.219 Due to
their superior water oxidation ability, IrO2 nanoparticles have
been extensively used in photoelectrocatalysts and in photo-
electrochemical cells. Nanostructured IrO2 films have been shown
to oxidize water with overpotentials as little as 0.2 V.287,288
The IrO2 particles are also often sensitized with Ru dyes89,244,289

This journal is c The Royal Society of Chemistry 2013 Chem. Soc. Rev., 2013, 42, 2294--2320 2311

Chem Soc Rev
or immobilized in Nafion membranes290 or linked to semi-
conductor nanoparticles.136,291,292 The charge transfer
dynamics involving IrO2, Ru dyes and TiO2 have been studied
separately by Kamat and Mallouk. Hole injection from a Ru dye
into IrO2 particles proceeds on the 2 ms timescale.293 From
TiO2 the hole transfer is considerably faster and occurs with a
rate constant of 6  105 s1.273
3d metal oxides
Limited availability and high prices prevent the large scale
application of noble metals for water electrolysis. This problem
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
is avoided with cocatalysts based on abundant elements. The
hope is to compensate for the lower per-atom activities of these
catalysts with the larger specific surface area of nanostructured
photoelectrodes. Nickel is one of the earliest abundant nano-
particle cocatalysts employed for water splitting, especially in
combination with large band gap semiconductor catalysts,
such as the tantalates, niobates, and titanates.37,44,294300 The
metal is usually applied by soaking the semiconductor with
nickel nitrate solution, followed by hydrogen treatment at
elevated temperature, and a subsequent oxygen treatment at
130 1C. The oxidation step converts some of the Ni metal into
NiO, which is necessary to achieve complete water photo-
electrolysis. High resolution electron microscopy established
that the NiO forms a shell around the Ni particles,301 which is
believed to function as an oxygen barrier, in analogy to the
Cr2O3 shell in the Rh/Cr2O3 system discussed above. The
oxidation step can be omitted if the Ni is used as a photo-
cathode material in a cell. According to recent work by Lewis
group electroplated NiMo alloy particles on silicon photo-
cathodes are nearly as active for the hydrogen evolution reac-
tion as e-beam deposited platinum.254 The performance also
depends strongly on the nanomaterial amount and deposition
method (Fig. 30). Among water oxidation catalysts, nano-Mn2O3
has recently emerged as a promising but cheap catalyst. After
electrodeposition onto a polished glassy carbon electrode,
Mn2O3 is active for both the oxygen evolution and oxygen
Fig. 30 Silicon microwires with cocatalysts deposited on top (A, C) and base (B, D) for the indicated times: (A, B) Ni 5s; (C, D) Ni 1s. Right: electrochemical polarization
curves during proton reduction. Reproduced with permission from ref. 254. Copyright 2011, Royal Society of Chemistry.
2312 Chem. Soc. Rev., 2013, 42, 2294--2320
View Article Online
Review Article
reduction reaction (ORR), with overpotentials of only 130 mV
higher than those for Ir, Rh, and Pt electrodes.302 The superior
function of the film may stem from its chemical relationship to
the oxygen evolving CaMn4Ox cluster in photosystem II. A
dierent manganese oxide, Mn3O4, also proved to be eective
for water oxidation. When nanocrystals of it were integrated
into a Rh/Cr2O3GaN:ZnO photocatalyst, an increase of the
overall water splitting activity was observed.282
Heinz Freis group recently reported that nanocrystals
of MnO2 inside of a mesoporous SiO2 matrix catalyzed O2
evolution after sensitization with a Ru bipyridyl complex and
in the presence of sodium persulfate.222 A similar reactivity
was observed for Co3O4 and with IrO2 nanocrystals,303 after
sensitization with a single Cr center.304 Lastly, we mention the
recent observation of photocatalytic properties for nanocrystals
of MnO2 (see above).
Co-Pi and Co-oxides
One of the most popular water oxidation catalysts is the Co-Pi
catalyst, invented by Dan Noceras group in 2008.305 The
amorphous material contains Co(III) ions, oxide and phosphate,
and films of it can electroxidize water in neutral phosphate
buer solution at 0.41 V overpotential (measured at 1 mA
cm2). In connection with a triple junction amorphous silicon
photovoltaic cell, and a NiMoZn water reduction catalyst, Co-Pi
films accomplish visible light driven overall water splitting with
up to 4.7% energy eciency. The device is known as the
artificial leaf.10 In the last three years, films of the Co-Pi
catalyst have been used to enhance the performance of Fe2O3
and ZnO photoelectrodes (for details see respective sections
above). But other cobalt oxides also have shown reasonable
activity for water oxidation.265 Don Tilleys group recently
reported on the use of Co, CoO, and Co3O4 nanoparticle
catalysts for electrochemical water oxidation.306 Styrings group
demonstrated photocatalytic water oxidation with Ru dye sensi-
tized Co oxide nanoparticles in sodium persulfate solution.307
Mullins group showed that the photocurrent of N-doped TiO2
This journal is c The Royal Society of Chemistry 2013

Review Article
nanorod arrays increased significantly upon addition of a cobalt
oxide layer.103
Metal sulfides
Molybdenum sulfide, MoS2, has recently emerged as an active HER
catalyst.308 Thin films of the material on Au support proton
reduction current densities of 0.3 mA cm3 at 0.17 V over-
potential. As the active sites of the catalyst are located at the edge
sites of the MoS2 sheets, the activity increases upon nanostructur-
ing the material. The catalytic properties are preserved in nano-WS2
and can be enhanced by addition of cobalt ions which bind to the
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
cluster edge sites.309 When MoS2 nanocrystals obtained by
chemical exfoliation are combined with CdSe nanoribbons, the
photocatalytic H2 evolution rate increases by up to 3.7 times.241 Ib
Chorkendors group recently employed the cuboidal Mo3S4 cluster
anion as a HER cocatalyst on p-Si pillar photocathode.310 Under
red-light illumination (l > 620 nm, 28.3 mW cm2) this configu-
ration supported photocurrents of up to 14 mA cm2, at 0.3 V
applied potential (in 1.0 M aqueous HClO4). The cocatalyst reduced
the proton reduction overpotential by 0.7 V.
Nanoparticles for plasmonic eects
A rather novel application of nanoparticles concerns their
surface plasmon resonance to boost visible light absorption
of photoelectrodes.311 The enhancement is thought to arise
from increased electrical field strength near the particle, and
due to enhanced light scattering throughout the photoelec-
trode. In 2010, Augustynskis group reported 60% enhanced
photocurrents from WO3 film electrodes that contained a silver
nanoparticle underlayer. The eect was attributed to improved
light absorption by WO3 caused by a combination of optical
(nonresonant light scattering and light reflection) and plasmo-
nic (resonant light scattering and electromagnetic field
enhancement) eects from the silver nanoparticles.153 Recent
results by Norskovs group on TiO2 films on FTO, with a Au
nanoparticle layer in between, show that increased photocurrents
can at least in part be due to increased electronic contact resulting
from the higher surface area of the gold nanoparticle layer.312 To
test plasmonic enhancement in Fe2O3 photoanodes, Peidong
Yangs group fabricated a thin Fe2O3 film on top of gold nanopillar
arrays (Fig. 31).313 The pillar electrodes gave 40% increased
Fig. 31 SEM images of Fe2O3 film on gold nanopillar array. Reproduced with
permission from ref. 313. Copyright 2012, American Chemical Society.
This journal is c The Royal Society of Chemistry 2013
View Article Online
Chem Soc Rev
photocurrents (at 1.5 V vs. RHE in 1 M NaOH) compared to films
on a flat gold substrate. Weak traces of the plasmon resonance in
the IPCE curves indicated that the enhancement involved both
surface plasmon resonances and photonic-mode light trapping in
the nanostructured films. The eect of improved charge transport
resulting from the increase in Fe2O3Au and Fe2O3electrolyte
interface was assumed as small.
In contrast, Gratzels group reported that 50 nm Au particles
embedded into a Fe2O3 photoanode film had no impact on the
photocurrent. Reduced photocurrent was observed for a Fe2O3
film with 50 nm Au particles on top of the film. This was
attributed to higher e/h recombination.314 However, after sub-
tracting this eect from the photocurrent spectrum, a small
enhancement in the 400600 nm region could be seen, that was
interpreted as a result of gold surface plasmon resonance.
Overall, these findings suggest that surface plasmon reso-
nances can impact light absorption and photocurrent genera-
tion by photoelectrodes, but that insulating coatings need to be
employed to suppress enhanced charge trapping, transfer, and
recombination at the metalsemiconductor interface.
Conclusion and outlook
The last decade has seen an increase in research activity
on nanostructured photoelectrochemical and photocatalytic
systems for solar water splitting. The nanoscaling approach
has been shown to particularly improve the performance of
metal oxide photoanode materials with low carrier mobility and
with short excited state lifetimes. For electrocatalysts, nanos-
caling has increased the electroactive surface area and allowed
for more ecient materials use. For selected metal chalcogenides,
the quantum size eect has been useful for controlling interfacial
charge transport, and in promoting photocatalytic proton
reduction. These examples clearly show the benefits of nano-
scaling and highlight its potential for the development of
improved photoelectrode materials.
At the same time it is apparent that the performance of a
material is ultimately defined by the intrinsic materials para-
meters, i.e. by its chemical composition and structure. For
many metal oxides, low electrical conductivity, high defect
concentrations, and short excited state lifetimes are important
problems that cannot be solved through nanoscaling alone.
New materials with ternary and quaternary compositions are
required here, to avoid these shortcomings. Even though
inexpensive nanomaterials with low proton reduction overpo-
tentials are now available (e.g. NiMo alloy), there is still a need
for more active water oxidation catalysts, based on earth
abundant elements. Here, nanostructuring can be a significant
asset in increasing the electroactive area, or in preventing back
reactions, as in the case of Cr2O3-coated Rh nanoparticles. Also,
due to the corrosive nature of water electrolysis, material
stability is a limiting issue that aects all non-oxide materials.
Nanostructured coatings with inert materials have been shown
to reduce corrosion, but not to eliminate it fully. Corrosion
must be considered as an important parameter in devising new
photocatalysts for water splitting. This particularly applies to
Chem. Soc. Rev., 2013, 42, 2294--2320 2313

Chem Soc Rev
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
Fig. 32 Chemical potential (A) versus kinetic (B) rectification. Adapted from
ref. 54. Copyright 2008, International Association of Hydrogen Energy.
assemblies that utilize non-oxide materials, or molecular com-
ponents as linkers, modifiers, or catalysts.
Some drawbacks of these materials, e.g. low electrical con-
ductivity, may be circumvented with suspended photocatalysts,
which are nanoscale in all three dimensions, and combine
photoanode and cathode materials in close proximity. Such
integrated photochemical diodes might have the additional
advantage over photoelectrochemical cells, that they would be
easier to manufacture on a large scale, and be less costly in
their operation.3 However, the key limitation of these systems is
insucient rectification. Since space charge layers are not
eective at the nanoscale, new ways of separating charge are
needed. One possibility are charge-selective interfaces that
could support kinetic rectification a concept that was
formulated by Helmut Tributsch in 2008,54 in analogy to
natural photosynthesis (Fig. 32). Alternatively, it might be
feasible to increase the doping levels of nanocrystals, to
improve the eectiveness of space charge layers on the nano-
scale, although that again might come at the expense of
increased recombination. These highly integrated systems
would also require higher precision in their synthesis/assembly,
especially if several components are involved. This includes
better control of interfaces, so that the energy loss processes
arising from the increased interfacial areas can be minimized.
The development of surface treatment techniques, e.g. the use of
chemical modifiers (Ga, Zn, Al ions), to reduce defects and
improve carrier lifetimes will be essential for this purpose.
Based on the desired function, it is unlikely that an ecient
solar energy driven photocatalytic water splitting system will
exceed the complexity of a microprocessor. It is just that the
methods for making and integrating the microscale compo-
nents at the necessary precision level are not available yet. To
make them available in the near future should be one of the
goals of scientists today. This will require not only more
systematic eorts in materials preparation, but also in
characterization and testing. For example, measurements of
the photoelectrochemical/photocatalytic performance should
follow established standards.315 This challenge is particularly
significant for nanomaterials, which show large property varia-
tions based on their morphology, surface termination, and
surface charge. Such factors must be considered in the analysis.
2314 Chem. Soc. Rev., 2013, 42, 2294--2320
View Article Online
Review Article
Acknowledgements
This material is based upon work supported by the National
Science Foundation under CHE 1152250 and CBET
1133099). Any opinions, findings, and conclusions or recom-
mendations expressed in this material are those of the author
and do not necessarily reflect the views of the National Science
Foundation. The author thanks Research Corporation for
Science Advancement for a Scialog award, and Michael A.
Holmes and Erwin M. Sabio for artwork.
References
1 N. S. Lewis and D. G. Nocera, Proc. Natl. Acad. Sci. U. S. A.,
2006, 103, 1572915735.
2 N. S. Lewis, G. Crabtree, A. J. Nozik, M. R. Wasielewski and
A. P. Alivisatos, Basic Research Needs for Solar Energy
Utilization, Department of Energy, 2005.
3 D. T. Inc., 2009.
4 M. G. Walter, E. L. Warren, J. R. McKone, S. W. Boettcher,
Q. X. Mi, E. A. Santori and N. S. Lewis, Chem. Rev., 2010,
110, 64466473.
5 O. Khaselev and J. A. Turner, Science, 1998, 280, 425427.
6 G. Peharz, F. Dimroth and U. Wittstadt, Int. J. Hydrogen
Energy, 2007, 32, 32483252.
7 J. R. Bolton, S. J. Strickler and J. S. Connolly, Nature, 1985,
316, 495500.
8 O. K. Varghese and C. A. Grimes, Sol. Energy Mater. Sol.
Cells, 2008, 92, 374384.
9 D. G. Nocera, Acc. Chem. Res., 2012, 45, 767776.
10 S. Y. Reece, J. A. Hamel, K. Sung, T. D. Jarvi, A. J. Esswein,
J. J. H. Pijpers and D. G. Nocera, Science, 2011, 334, 645648.
11 U. A. Joshi, A. Palasyuk, D. Arney and P. A. Maggard,
J. Phys. Chem. Lett., 2010, 1, 27192726.
12 M. Woodhouse and B. A. Parkinson, Chem. Mater., 2008,
20, 24952502.
13 M. Woodhouse, G. S. Herman and B. A. Parkinson, Chem.
Mater., 2005, 17, 43184324.
14 M. Woodhouse and B. A. Parkinson, Chem. Soc. Rev., 2009,
38, 197210.
15 T. F. Jaramillo, S. H. Baeck, A. Kleiman-Shwarsctein,
K. S. Choi, G. D. Stucky and E. W. McFarland, J. Comb.
Chem., 2005, 7, 264271.
16 J. E. Katz, T. R. Gingrich, E. A. Santori and N. S. Lewis,
Energy Environ. Sci., 2009, 2, 103112.
17 R. Dingle, W. Wiegmann and C. H. Henry, Phys. Rev. Lett.,
1974, 33, 827830.
18 A. Henglein, Ber. BunsenGes., 1982, 86, 301305.
19 A. Fojtik, H. Weller, U. Koch and A. Henglein, Ber. Bunsen
Ges., 1984, 88, 969977.
20 L. E. Brus, J. Chem. Phys., 1983, 79, 55665571.
21 W. H. Brattain and C. G. B. Garrett, Bell Syst. Tech. J., 1955,
34, 129176.
22 H. Gerischer, J. Electrochem. Soc., 1966, 113, 11741182.
23 Y. V. Pleskov and Y. Y. Gurevich, Semiconductor photo-
electrochemistry, Consultants Bureau, New York, 1986.
This journal is c The Royal Society of Chemistry 2013

Review Article
24 R. A. Marcus, Annu. Rev. Phys. Chem., 1964, 15, 155196.
25 A. Fujishima and K. Honda, Bull. Chem. Soc. Jpn., 1971, 44,
11481150.
26 A. Fujishima and K. Honda, Nature, 1972, 238, 3738.
27 A. J. Bard, J. Photochem., 1979, 10, 5975.
28 A. J. Nozik, Annu. Rev. Phys. Chem., 1978, 29, 189222.
29 A. J. Bard and M. A. Fox, Acc. Chem. Res., 1995, 28, 141145.
30 A. Mills and S. LeHunte, J. Photochem. Photobiol., A, 1997,
108, 135.
31 R. Memming, in Electron Transfer I, ed. J. Mattay, 1994,
vol. 169, pp. 105181.
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
32 A. Mills and G. Porter, J. Chem. Soc., Faraday Trans. 1, 1982,
78, 36593669.
33 D. Duonghong, E. Borgarello and M. Gratzel, J. Am. Chem.
Soc., 1981, 103, 46854690.
34 K. Rajeshwar, J. Appl. Electrochem., 2007, 37, 765787.
35 K. Maeda, J. Photochem. Photobiol., C, 2011, 12, 237268.
36 K. Maeda and K. Domen, J. Phys. Chem. C, 2007, 111,
78517861.
37 A. Kudo and Y. Miseki, Chem. Soc. Rev., 2009, 38, 253278.
38 R. Abe, J. Photochem. Photobiol., C, 2010, 11, 179209.
39 F. E. Osterloh, Chem. Mater., 2008, 20, 3554.
40 F. E. Osterloh and B. A. Parkinson, MRS Bull., 2011, 36, 1722.
41 T. Yokoi, J. Sakuma, K. Maeda, K. Domen, T. Tatsumi and
J. N. Kondo, Phys. Chem. Chem. Phys., 2011, 13, 25632570.
42 K. Saito, K. Koga and A. Kudo, Dalton Trans., 2011, 40,
39093913.
43 S. C. Yan, L. J. Wan, Z. S. Li and Z. G. Zou, Chem. Commun.,
2011, 47, 56325634.
44 H. Kato, K. Asakura and A. Kudo, J. Am. Chem. Soc., 2003,
125, 30823089.
45 T. K. Townsend, N. D. Browning and F. E. Osterloh, ACS
Nano, 2012, 6, 74207426.
46 S. D. Tilley, M. Cornuz, K. Sivula and M. Gratzel, Angew.
Chem., Int. Ed., 2010, 49, 64056408.
47 M. P. Dare-Edwards, J. B. Goodenough, A. Hamnett and
P. R. Trevellick, J. Chem. Soc., Faraday. Trans. 1, 1983, 79,
20272041.
48 B. Oregan and M. Gratzel, Nature, 1991, 353, 737740.
49 S. W. Keller, S. A. Johnson, E. S. Brigham, E. H. Yonemoto and
T. E. Mallouk, J. Am. Chem. Soc., 1995, 117, 1287912880.
50 D. M. Kaschak, J. T. Lean, C. C. Waraksa, G. B. Saupe,
H. Usami and T. E. Mallouk, J. Am. Chem. Soc., 1999, 121,
34353445.
51 L. I. Berger, in CRC Handbook of Chemistry and Physics,
ed. D. R. Lide, CRC Press/Taylor and Francis, Boca Raton,
FL, 88 (Internet Version 2008), 2008.
52 M. N. Huda, M. M. Al-Jassim and J. A. Turner, J. Renewable
Sustainable Energy, 2011, 3, 053101.
53 P. A. Cox, Transition metal oxides: an introduction to their
electronic structure and properties, Clarendon Press; Oxford
University Press, Oxford, New York, 2010.
54 H. Tributsch, Int. J. Hydrogen Energy, 2008, 33, 59115930.
55 L. I. Berger, in CRC Handbook of Chemistry and Physics, ed.
D. R. Lide, CRC Press/Taylor and Francis, Boca Raton, FL,
88 (Internet Version 2008), 2008.
This journal is c The Royal Society of Chemistry 2013
View Article Online
Chem Soc Rev
56 A. Polman and H. A. Atwater, Nat. Mater., 2012, 11,
174177.
57 A. D. Yoe, Adv. Phys., 2001, 50, 1208.
58 P. Salvador, J. Phys. Chem. B, 2001, 105, 61286141.
59 H. Gerischer, Electrochim. Acta, 1990, 35, 16771699.
60 P. A. Sant and P. V. Kamat, Phys. Chem. Chem. Phys., 2002,
4, 198203.
61 I. Robel, M. Kuno and P. V. Kamat, J. Am. Chem. Soc, 2007,
129, 41364137.
62 M. A. Holmes, T. K. Townsend and F. E. Osterloh, Chem.
Commun., 2012, 48, 371373.
63 M. A. Butler and D. S. Ginley, J. Electrochem. Soc., 1978, 125,
228232.
64 P. Singh, R. Singh, R. Gale, K. Rajeshwar and J. Dubow,
J. Appl. Phys., 1980, 51, 62866291.
65 L. Kavan, M. Gratzel, S. E. Gilbert, C. Klemenz and
H. J. Scheel, J. Am. Chem. Soc., 1996, 118, 67166723.
66 L. M. Peter, K. G. U. Wijayantha, D. J. Riley and
J. P. Waggett, J. Phys. Chem. B, 2003, 107, 83788381.
67 V. Chakrapani, K. Tvrdy and P. V. Kamat, J. Am. Chem. Soc.,
2010, 132, 12281229.
68 A. J. Cowan, J. W. Tang, W. H. Leng, J. R. Durrant and
D. R. Klug, J. Phys. Chem. C, 2010, 114, 42084214.
69 J. W. Tang, J. R. Durrant and D. R. Klug, J. Am. Chem. Soc.,
2008, 130, 1388513891.
70 J. B. Sambur, T. Novet and B. A. Parkinson, Science, 2010,
330, 6366.
71 O. E. Semonin, J. M. Luther, S. Choi, H. Y. Chen,
J. B. Gao, A. J. Nozik and M. C. Beard, Science, 2011, 334,
15301533.
72 A. J. Nozik, Physica E, 2002, 14, 115120.
73 E. Yablonovitch, D. L. Allara, C. C. Chang, T. Gmitter and
T. B. Bright, Phy. Rev. Lett., 1986, 57, 249252.
74 N. S. Lewis, Inorg. Chem., 2005, 44, 69006911.
75 G. Hodes, I. D. J. Howell and L. M. Peter, J. Electrochem.
Soc., 1992, 139, 31363140.
76 I. Cesar, K. Sivula, A. Kay, R. Zboril and M. Graetzel, J. Phys.
Chem. C, 2009, 113, 772782.
77 A. Hagfeldt and M. Gratzel, Chem. Rev., 1995, 95, 4968.
78 B. Oregan, J. Moser, M. Anderson and M. Gratzel, J. Phys.
Chem., 1990, 94, 87208726.
79 L. Dloczik, O. Ileperuma, I. Lauermann, L. M. Peter,
E. A. Ponomarev, G. Redmond, N. J. Shaw and
I. Uhlendorf, J. Phys. Chem. B, 1997, 101, 1028110289.
80 N. C. Giebink, G. P. Wiederrecht, M. R. Wasielewski and
S. R. Forrest, Phys. Rev. B: Condens. Matter Mater. Phys.,
2011, 83, 195326.
81 P. Wu rfel, Wiley-VCH, Weinheim, 2005, p. 142.
82 S. Sodergren, A. Hagfeldt, J. Olsson and S. E. Lindquist,
J. Phys. Chem., 1994, 98, 55525556.
83 A. Hagfeldt, U. Bjorksten and S. E. Lindquist, Sol. Energ.
Mater. Sol. Cells, 1992, 27, 293304.
84 J. Bisquert and V. S. Vikhrenko, J. Phys. Chem. B, 2004, 108,
23132322.
85 A. Hagfeldt, H. Lindstrom, S. Sodergren and S. E. Lindquist,
J. Electroanal. Chem., 1995, 381, 3946.
Chem. Soc. Rev., 2013, 42, 2294--2320 2315

Chem Soc Rev
86 B. Abeles, P. Sheng, M. D. Coutts and Y. Arie, Adv. Phys.,
1975, 24, 407461.
87 R. H. Terrill, T. A. Postlethwaite, C. H. Chen, C. D. Poon,
A. Terzis, A. D. Chen, J. E. Hutchison, M. R. Clark,
G. Wignall, J. D. Londono, R. Superfine, M. Falvo,
C. S. Johnson, E. T. Samulski and R. W. Murray, J. Am.
Chem. Soc., 1995, 117, 1253712548.
88 L. Vayssieres, On solar hydrogen & nanotechnology, John
Wiley & Sons (Asia), Singapore, Hoboken, NJ, 2009.
89 P. G. Hoertz and T. E. Mallouk, Inorg. Chem., 2005, 44,
68286840.
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
90 J. F. Zhu and M. Zach, Curr. Opin. Colloid Interface Sci.,
2009, 14, 260269.
91 P. V. Kamat, J. Phys. Chem. C, 2007, 111, 28342860.
92 P. V. Kamat and N. M. Dimitrijevic, Sol. Energy, 1990, 44,
8398.
93 P. V. Kamat, K. Tvrdy, D. R. Baker and J. G. Radich, Chem.
Rev., 2010, 110, 66646688.
94 J. Z. Zhang, MRS Bull., 2011, 36, 4855.
95 J. M. Foley, M. J. Price, J. I. Feldblyum and S. Maldonado,
Energy Environ. Sci., 2012, 5, 52035220.
96 W. Jaegermann and H. Tributsch, Prog. Surf. Sci., 1988, 29,
1167.
97 A. Wahl and J. Augustynski, J. Phys. Chem. B, 1998, 102,
78207828.
98 X. J. Feng, K. Zhu, A. J. Frank, C. A. Grimes and
T. E. Mallouk, Angew. Chem., Int. Ed., 2012, 51, 27272730.
99 S. Hoang, S. W. Guo, N. T. Hahn, A. J. Bard and
C. B. Mullins, Nano Lett., 2012, 12, 2632.
100 I. S. Cho, Z. B. Chen, A. J. Forman, D. R. Kim, P. M. Rao,
T. F. Jaramillo and X. L. Zheng, Nano Lett., 2011, 11, 49784984.
101 K. Shankar, J. I. Basham, N. K. Allam, O. K. Varghese,
G. K. Mor, X. J. Feng, M. Paulose, J. A. Seabold, K. S. Choi
and C. A. Grimes, J. Phys. Chem. C, 2009, 113, 63276359.
102 G. K. Mor, K. Shankar, M. Paulose, O. K. Varghese and
C. A. Grimes, Nano Lett., 2005, 5, 191195.
103 S. P. Berglund, D. W. Flaherty, N. T. Hahn, A. J. Bard and
C. B. Mullins, J. Phys. Chem. C, 2011, 115, 37943802.
104 S. Hoang, S. P. Berglund, N. T. Hahn, A. J. Bard and
C. B. Mullins, J. Am. Chem. Soc., 2012, 134, 36593662.
105 G. K. Mor, O. K. Varghese, R. H. T. Wilke, S. Sharma,
K. Shankar, T. J. Latempa, K. S. Choi and C. A. Grimes,
Nano Lett., 2008, 8, 19061911.
106 J. R. Jennings, A. Ghicov, L. M. Peter, P. Schmuki and
A. B. Walker, J. Am. Chem. Soc., 2008, 130, 1336413372.
107 W. H. Leng, P. R. F. Barnes, M. Juozapavicius, B. C. ORegan
and J. R. Durrant, J. Phys. Chem. Lett., 2010, 1, 967972.
108 L. Q. Jing, J. Zhou, J. R. Durrant, J. W. Tang, D. N. Liu and
H. G. Fu, Energy Environ. Sci., 2012, 5, 65526558.
109 J. Zhang, J. H. Bang, C. C. Tang and P. V. Kamat, ACS Nano,
2010, 4, 387395.
110 A. Kongkanand, R. M. Dominguez and P. V. Kamat, Nano
Lett., 2007, 7, 676680.
111 A. J. Cowan, C. J. Barnett, S. R. Pendlebury, M. Barroso,
K. Sivula, M. Gratzel, J. R. Durrant and D. R. Klug, J. Am.
Chem. Soc., 2011, 133, 1013410140.
2316 Chem. Soc. Rev., 2013, 42, 2294--2320
View Article Online
Review Article
112 H. Durr, S. Bossmann and A. Beuerlein, J. Photochem.
Photobiol., A, 1993, 73, 233245.
113 J. Kiwi, E. Borgarello, E. Pelizzetti, M. Visca and M. Gratzel,
Angew. Chem., Int. Ed., 1980, 19, 646648.
114 M. Gratzel, Acc. Chem. Res., 1981, 14, 376384.
115 D. Duonghong and M. Gratzel, J. Chem. Soc., Chem. Com-
mun., 1984, 15971599.
116 J. Kiwi and M. Gratzel, J. Phys. Chem., 1984, 88, 13021307.
117 B. Gu, J. Kiwi and M. Gratzel, New J. Chem., 1985, 9,
539543.
118 M. R. Allen, A. Thibert, E. M. Sabio, N. D. Browning,
D. S. Larsen and F. E. Osterloh, Chem. Mater., 2010, 22,
12201228.
119 X. B. Chen, L. Liu, P. Y. Yu and S. S. Mao, Science, 2011,
331, 746750.
120 A. Danon, K. Bhattacharyya, B. K. Vijayan, J. L. Lu,
D. J. Sauter, K. A. Gray, P. C. Stair and E. Weitz, ACS Catal.,
2012, 2, 4549.
121 E. M. P. Steinmiller and K. S. Choi, Proc. Natl. Acad. Sci.
U. S. A., 2009, 106, 2063320636.
122 K. Sun, K. Madsen, P. Andersen, W. N. Bao, Z. L. Sun and
D. L. Wang, Nanotechnology, 2012, 23, 194013.
123 A. Wolcott, W. A. Smith, T. R. Kuykendall, Y. P. Zhao and
J. Z. Zhang, Adv. Funct. Mater., 2009, 19, 18491856.
124 K. Domen, A. Kudo, M. Shibata, A. Tanaka, K. Maruya and
T. Onishi, J. Chem. Soc., Chem. Commun., 1986, 17061707.
125 O. C. Compton, C. H. Mullet, S. Chiang and F. E. Osterloh,
J. Phys. Chem. C, 2008, 112, 62026208.
126 T. K. Townsend, E. M. Sabio, N. D. Browning and
F. E. Osterloh, ChemSusChem, 2011, 4, 185190.
127 M. C. Sarahan, E. C. Carroll, M. Allen, D. S. Larsen,
N. D. Browning and F. E. Osterloh, J. Solid State Chem.,
2008, 181, 16811686.
128 O. C. Compton and F. E. Osterloh, J. Phys. Chem. C, 2009,
113, 479485.
129 O. C. Compton, E. C. Carroll, J. Y. Kim, D. S. Larsen and
F. E. Osterloh, J. Phys. Chem. C, 2007, 111, 1458914592.
130 E. M. Sabio, M. Chi, N. D. Browning and F. E. Osterloh,
Langmuir, 2010, 26, 72547261.
131 E. C. Carroll, O. C. Compton, D. Madsen, D. S. Larsen and
F. E. Osterloh, J. Phys. Chem. C, 2008, 112, 23942403.
132 E. M. Sabio, R. L. Chamousis, N. D. Browning and
F. E. Osterloh, J. Phys. Chem. C, 2012, 116, 31613170;
J. Phys. Chem. C, 2012, 116, 19051.
133 Y. Ebina, T. Sasaki, M. Harada and M. Watanabe, Chem.
Mater., 2002, 14, 43904395.
134 K. Maeda and T. E. Mallouk, J. Mater. Chem., 2009, 19,
48134818.
135 Y. Okamoto, S. Ida, J. Hyodo, H. Hagiwara and T. Ishihara,
J. Am. Chem. Soc., 2011, 133, 1803418037.
136 W. J. Youngblood, S. H. A. Lee, K. Maeda and
T. E. Mallouk, Acc. Chem. Res., 2009, 42, 19661973.
137 Y. I. Kim, S. Salim, M. J. Huq and T. E. Mallouk, J. Am.
Chem. Soc., 1991, 113, 95619563.
138 Y. I. Kim, S. J. Atherton, E. S. Brigham and T. E. Mallouk,
J. Phys. Chem., 1993, 97, 1180211810.
This journal is c The Royal Society of Chemistry 2013

Review Article
139 K. Maeda, M. Eguchi, W. J. Youngblood and T. E. Mallouk,
Chem. Mater., 2008, 20, 67706778.
140 K. Maeda, M. Eguchi, W. J. Youngblood and T. E. Mallouk,
Chem. Mater., 2009, 21, 36113617.
141 H. Zhou, E. M. Sabio, T. K. Townsend, T. Fan, D. Zhang
and F. E. Osterloh, Chem. Mater., 2010, 22, 33623368.
142 K. Maeda, M. Eguchi, S. H. A. Lee, W. J. Youngblood, H. Hata
and T. E. Mallouk, J. Phys. Chem. C, 2009, 113, 79627969.
143 G. Hodes, D. Cahen and J. Manassen, Nature, 1976, 260,
312313.
144 B. Marsen, E. L. Miller, D. Paluselli and R. E. Rocheleau,
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
Int. J. Hydrogen Energy, 2007, 32, 31103115.
145 C. Santato, M. Ulmann and J. Augustynski, Adv. Mater.,
2001, 13, 511514.
146 C. Santato, M. Ulmann and J. Augustynski, J. Phys. Chem. B,
2001, 105, 936940.
147 R. Solarska, R. Jurczakowski and J. Augustynski, Nanoscale,
2012, 4, 15531556.
148 R. Solarska, B. D. Alexander, A. Braun, R. Jurczakowski,
G. Fortunato, M. Stiefel, T. Graule and J. Augustynski,
Electrochim. Acta, 2010, 55, 77807787.
149 V. Chakrapani, J. Thangala and M. K. Sunkara, Int. J.
Hydrogen Energy, 2009, 34, 90509059.
150 S. H. Baeck, K. S. Choi, T. F. Jaramillo, G. D. Stucky and
E. W. McFarland, Adv. Mater., 2003, 15, 12691273.
151 D. D. Qin, C. L. Tao, S. A. Friesen, T. H. Wang,
O. K. Varghese, N. Z. Bao, Z. Y. Yang, T. E. Mallouk and
C. A. Grimes, Chem. Commun., 2012, 48, 729731.
152 F. Amano, D. Li and B. Ohtani, J. Electrochem. Soc., 2011,
158, K42K46.
153 R. Solarska, A. Krolikowska and J. Augustynski, Angew.
Chem., Int. Ed., 2010, 49, 79807983.
154 F. M. Pesci, A. J. Cowan, B. D. Alexander, J. R. Durrant and
D. R. Klug, J. Phys. Chem. Lett., 2011, 2, 19001903.
155 J. A. Seabold and K. S. Choi, Chem. Mater., 2011, 23,
11051112.
156 J. R. Darwent and A. Mills, J. Chem. Soc., Faraday. Trans. 2,
1982, 78, 359367.
157 M. Waller, T. K. Townsend, J. Zhao, E. M. Sabio,
R. L. Chamousis, N. D. Browning and F. E. Osterloh, Chem.
Mater., 2012, 24, 698704.
158 A. Paracchino, V. Laporte, K. Sivula, M. Graetzel and
E. Thimsen, Nat. Mater., 2011, 10, 456461.
159 C. M. McShane and K. S. Choi, J. Am. Chem. Soc., 2009, 131,
25612569.
160 L. Chen, S. Shet, H. W. Tang, H. L. Wang, T. Deutsch,
Y. F. Yan, J. Turner and M. Al-Jassim, J. Mater. Chem., 2010,
20, 69626967.
161 A. Kudo, K. Ueda, H. Kato and I. Mikami, Catal. Lett., 1998,
53, 229230.
162 S. K. Pilli, T. E. Furtak, L. D. Brown, T. G. Deutsch,
J. A. Turner and A. M. Herring, Energy Environ. Sci., 2011,
4, 50285034.
163 S. K. Pilli, T. G. Deutsch, T. E. Furtak, J. A. Turner,
L. D. Brown and A. M. Herring, Phys. Chem. Chem. Phys.,
2012, 14, 70327039.
This journal is c The Royal Society of Chemistry 2013
View Article Online
Chem Soc Rev
164 J. A. Seabold and K. S. Choi, J. Am. Chem. Soc., 2012, 134,
21862192.
165 F. F. Abdi and R. van de Krol, J. Phys. Chem. C, 2012, 116,
93989404.
166 K. L. Hardee and A. J. Bard, J. Electrochem. Soc., 1976, 123,
10241026.
167 J. H. Kennedy and K. W. Frese, J. Electrochem. Soc., 1978,
125, 709714.
168 K. Itoh and J. O. Bockris, J. Electrochem. Soc., 1984, 131,
12661271.
169 S. A. Majumder and S. U. M. Khan, Int. J. Hydrogen Energy,
1994, 19, 881887.
170 S. U. M. Khan and J. Akikusa, J. Phys. Chem. B, 1999, 103,
71847189.
171 C. J. Sartoretti, M. Ulmann, B. D. Alexander, J. Augustynski
and A. Weidenka, Chem. Phys. Lett., 2003, 376, 194200.
172 C. J. Sartoretti, B. D. Alexander, R. Solarska,
W. A. Rutkowska, J. Augustynski and R. Cerny, J. Phys.
Chem. B, 2005, 109, 1368513692.
173 A. Duret and M. Gratzel, J. Phys. Chem. B, 2005, 109,
1718417191.
174 A. Kay, I. Cesar and M. Gratzel, J. Am. Chem. Soc., 2006, 128,
1571415721.
175 F. Le Formal, M. Gratzel and K. Sivula, Adv. Funct. Mater.,
2010, 20, 10991107.
176 I. Cesar, A. Kay, J. A. G. Martinez and M. Gratzel, J. Am.
Chem. Soc., 2006, 128, 45824583.
177 R. H. Goncalves, B. H. R. Lima and E. R. Leite, J. Am. Chem.
Soc., 2011, 133, 60126019.
178 J. Brillet, M. Gratzel and K. Sivula, Nano Lett., 2010, 10,
41554160.
179 L. Vayssieres, C. Sathe, S. M. Butorin, D. K. Shuh,
J. Nordgren and J. H. Guo, Adv. Mater., 2005, 17, 23202323.
180 L. Vayssieres, N. Beermann, S. E. Lindquist and
A. Hagfeldt, Chem. Mater., 2001, 13, 233235.
181 B. D. Chernomordik, H. B. Russell, U. Cvelbar,
J. B. Jasinski, V. Kumar, T. Deutsch and M. K. Sunkara,
Nanotechnology, 2012, 23, 194009.
182 N. T. Hahn, H. C. Ye, D. W. Flaherty, A. J. Bard and
C. B. Mullins, ACS Nano, 2010, 4, 19771986.
183 K. Sivula, R. Zboril, F. Le Formal, R. Robert, A. Weidenka,
J. Tucek, J. Frydrych and M. Gratzel, J. Am. Chem. Soc.,
2010, 132, 74367444.
184 Y. J. Lin, S. Zhou, S. W. Sheehan and D. W. Wang, J. Am.
Chem. Soc., 2011, 133, 23982401.
185 H. Dotan, K. Sivula, M. Gratzel, A. Rothschild and
S. C. Warren, Energy Environ. Sci., 2011, 4, 958964.
186 C. Y. Cummings, F. Marken, L. M. Peter, A. A. Tahir and
K. G. U. Wijayantha, Chem. Commun., 2012, 48, 20272029.
187 K. J. McDonald and K. S. Choi, Chem. Mater., 2011, 23,
48634869.
188 R. L. Spray, K. J. McDonald and K.-S. Choi, J. Phys. Chem. C,
2011, 115, 34973506.
189 T. Hisatomi, H. Dotan, M. Stefik, K. Sivula, A. Rothschild,
M. Gratzel and N. Mathews, Adv. Mater., 2012, 24,
26992702.
Chem. Soc. Rev., 2013, 42, 2294--2320 2317

Chem Soc Rev
190 F. Le Formal, N. Tetreault, M. Cornuz, T. Moehl, M. Gratzel
and K. Sivula, Chem. Sci., 2011, 2, 737743.
191 C. E. Kendrick, H. P. Yoon, Y. A. Yuwen, G. D. Barber,
H. T. Shen, T. E. Mallouk, E. C. Dickey, T. S. Mayer and
J. M. Redwing, Appl. Phys. Lett., 2010, 97, 143108.
192 M. Barroso, A. J. Cowan, S. R. Pendlebury, M. Gratzel,
D. R. Klug and J. R. Durrant, J. Am. Chem. Soc., 2011, 133,
1486814871.
193 C. Y. Cummings, F. Marken, L. M. Peter, K. G. U.
Wijayantha and A. A. Tahir, J. Am. Chem. Soc., 2012, 134,
12281234.
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
194 L. M. Peter, K. G. U. Wijayantha and A. A. Tahir, Faraday
Discuss., 2012, 155, 309322.
195 B. Klahr, S. Gimenez, F. Fabregat-Santiago, J. Bisquert and
T. W. Hamann, Energy Environ. Sci., 2012, 5, 76267636.
196 S. R. Pendlebury, M. Barroso, A. J. Cowan, K. Sivula,
J. W. Tang, M. Gratzel, D. Klug and J. R. Durrant, Chem.
Commun., 2011, 47, 716718.
197 S. R. Pendlebury, A. J. Cowan, M. Barroso, K. Sivula,
J. H. Ye, M. Gratzel, D. R. Klug, J. W. Tang and
J. R. Durrant, Energy Environ. Sci., 2012, 5, 63046312.
198 D. K. Zhong, J. W. Sun, H. Inumaru and D. R. Gamelin,
J. Am. Chem. Soc., 2009, 131, 60866087.
199 D. K. Zhong, M. Cornuz, K. Sivula, M. Graetzel and
D. R. Gamelin, Energy Environ. Sci., 2011, 4, 17591764.
200 K. J. McDonald and K.-S. Choi, Chem. Mater., 2011, 23,
16861693.
201 B. M. Klahr and T. W. Hamann, Appl. Phys. Lett., 2011,
99, 063508.
202 A. Kleiman-Shwarsctein, Y. S. Hu, A. J. Forman,
G. D. Stucky and E. W. McFarland, J. Phys. Chem. C,
2008, 112, 1590015907.
203 Y. S. Hu, A. Kleiman-Shwarsctein, A. J. Forman, D. Hazen,
J. N. Park and E. W. McFarland, Chem. Mater., 2008, 20,
38033805.
204 D. D. Qin, C. L. Tao, S. I. In, Z. Y. Yang, T. E. Mallouk, N. Z. Bao
and C. A. Grimes, Energy Fuels, 2011, 25, 52575263.
205 Y. C. Ling, G. M. Wang, D. A. Wheeler, J. Z. Zhang and
Y. Li, Nano Lett., 2011, 11, 21192125.
206 H. W. Tang, M. A. Matin, H. L. Wang, T. Deutsch, M. Al-
Jassim, J. Turner and Y. F. Yan, J. Appl. Phys., 2011,
110, 123511.
207 H. W. Tang, W. J. Yin, M. A. Matin, H. L. Wang, T. Deutsch,
M. M. Al-Jassim, J. A. Turner and Y. F. Yan, J. Appl. Phys.,
2012, 111, 073502.
208 G. M. Wang, Y. C. Ling, D. A. Wheeler, K. E. N. George,
K. Horsley, C. Heske, J. Z. Zhang and Y. Li, Nano Lett.,
2011, 11, 35033509.
209 A. Bak, W. Choi and H. Park, Appl. Catal., B, 2011, 110,
207215.
210 Y. S. Hu, A. Kleiman-Shwarsctein, G. D. Stucky and
E. W. McFarland, Chem. Commun., 2009, 26522654.
211 J. Kiwi and M. Gratzel, J. Chem. Soc. Faraday. Trans. 1,
1987, 83, 11011108.
212 M. A. Gondal, A. Hameed, Z. H. Yamani and A. Suwaiyan,
Appl. Catal., A, 2004, 268, 159167.
2318 Chem. Soc. Rev., 2013, 42, 2294--2320
View Article Online
Review Article
213 T. K. Townsend, E. M. Sabio, N. D. Browning and
F. E. Osterloh, Energy Environ. Sci., 2011, 4, 42704275.
214 M. Yagi, E. Tomita and T. Kuwabara, J. Electroanal. Chem.,
2005, 579, 8388.
215 S. Hackwood, L. M. Schiavone, W. C. Dautremontsmith
and G. Beni, J. Electrochem. Soc., 1981, 128, 25692573.
216 T. Nakagawa, C. A. Beasley and R. W. Murray, J. Phys.
Chem. C, 2009, 113, 1295812961.
217 A. Harriman, J. M. Thomas and G. R. Millward, New J.
Chem., 1987, 11, 757.
218 A. Mills and T. Russell, J. Chem. Soc., Faraday Trans., 1991,
87, 12451250.
219 N. D. Morris and T. E. Mallouk, J. Am. Chem. Soc., 2002,
124, 1111411121.
220 F. A. Frame, T. K. Townsend, R. L. Chamousis, E. M. Sabio,
T. Dittrich, N. D. Browning and F. E. Osterloh, J. Am. Chem.
Soc., 2011, 133, 72647267.
221 H. P. Maruska and A. K. Ghosh, Sol. Energy, 1978, 20,
443458.
222 F. Jiao and H. Frei, Chem. Commun., 2010, 46, 29202922.
223 N. Sakai, Y. Ebina, K. Takada and T. Sasaki, J. Phys. Chem.
B, 2005, 109, 96519655.
224 K. Izawa, T. Yamada, U. Unal, S. Ida, O. Altuntasoglu,
M. Koinuma and Y. Matsumoto, J. Phys. Chem. B, 2006,
110, 46454650.
225 R. Brimblecombe, G. F. Swiegers, G. C. Dismukes and
L. Spiccia, Angew. Chem., Int. Ed., 2008, 47, 73357338.
226 R. K. Hocking, R. Brimblecombe, L. Y. Chang, A. Singh,
M. H. Cheah, C. Glover, W. H. Casey and L. Spiccia, Nat.
Chem., 2011, 3, 461466.
227 K. J. Young, Y. L. Gao and G. W. Brudvig, Aust. J. Chem.,
2011, 64, 12191226.
228 B. A. Pinaud, Z. B. Chen, D. N. Abram and T. F. Jaramillo,
J. Phys. Chem. C, 2011, 115, 1183011838.
229 Y. Q. Wang, Z. J. Zhang, Y. Zhu, Z. C. Li, R. Vajtai, L. J. Ci
and P. M. Ajayan, ACS Nano, 2008, 2, 14921496.
230 L. Yuliati, J. H. Yang, X. C. Wang, K. Maeda, T. Takata,
M. Antonietti and K. Domen, J. Mater. Chem., 2010, 20,
42954298.
231 Y. Q. Cong, H. S. Park, S. J. Wang, H. X. Dang, F. R. F. Fan,
C. B. Mullins and A. J. Bard, J. Phys. Chem. C, 2012, 116,
1454114550.
232 R. Vogel, P. Hoyer and H. Weller, J. Phys. Chem., 1994, 98,
31833188.
233 L. Kronik, N. Ashkenasy, M. Leibovitch, E. Fefer, S. Yoram,
S. Gorer and G. Hodes, J. Electrochem. Soc., 1998, 145,
17481755.
234 K. Tvrdy, P. A. Frantsuzov and P. V. Kamat, Proc. Natl. Acad.
Sci. U. S. A., 2011, 108, 2934.
235 X. B. Chen, S. H. Shen, L. J. Guo and S. S. Mao, Chem. Rev.,
2010, 110, 65036570.
236 J. F. Reber and K. Meier, J. Phys. Chem., 1984, 88,
59035913.
237 N. Buhler, K. Meier and J. F. Reber, J. Phys. Chem., 1984, 88,
32613268.
238 J. F. Reber and M. Rusek, J. Phys. Chem., 1986, 90, 824834.
This journal is c The Royal Society of Chemistry 2013

Review Article
239 S. Kambe, M. Fujii, T. Kawai, S. Kawai and F. Nakahara,
Chem. Phys. Lett., 1984, 109, 105109.
240 F. A. Frame, E. C. Carroll, D. S. Larsen, M. S. Sarahan,
N. D. Browning and F. E. Osterloh, Chem. Commun., 2008,
22062208.
241 F. A. Frame and F. E. Osterloh, J. Phys. Chem. C, 2010, 114,
1062810633.
242 A. Kongkanand, K. Tvrdy, K. Takechi, M. Kuno and
P. V. Kamat, J. Am. Chem. Soc., 2008, 130, 40074015.
243 I. Robel, V. Subramanian, M. Kuno and P. V. Kamat, J. Am.
Chem. Soc., 2006, 128, 23852393.
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
244 P. Hoyer and R. Konenkamp, Appl. Phys. Lett., 1995, 66,
349351.
245 J. Choi, S. Y. Ryu, W. Balcerski, T. K. Lee and
M. R. Homann, J. Mater. Chem., 2008, 18, 23712378.
246 A. M. Zaniewski, M. Loster, B. Sadtler, A. P. Alivisatos and
A. Zettl, Phys. Rev. B: Condens. Matter Mater. Phys., 2010,
82, 155311.
247 B. D. Yuhas, A. L. Smeigh, A. P. S. Samuel, Y. Shim, S. Bag,
A. P. Douvalis, M. R. Wasielewski and M. G. Kanatzidis,
J. Am. Chem. Soc., 2011, 133, 72527255.
248 S. W. Boettcher, E. L. Warren, M. C. Putnam, E. A. Santori,
D. Turner-Evans, M. D. Kelzenberg, M. G. Walter,
J. R. McKone, B. S. Brunschwig, H. A. Atwater and
N. S. Lewis, J. Am. Chem. Soc., 2011, 133, 12161219.
249 B. M. Kayes, H. A. Atwater and N. S. Lewis, J. Appl. Phys.,
2005, 97, 114302.
250 E. A. Santori, J. R. Maiolo, M. J. Bierman, N. C. Strandwitz,
M. D. Kelzenberg, B. S. Brunschwig, H. A. Atwater and
N. S. Lewis, Energy Environ. Sci., 2012, 5, 68676871.
251 K. Hagedorn, C. Forgacs, S. Collins and S. Maldonado,
J. Phys. Chem. C, 2010, 114, 1201012017.
252 M. D. Kelzenberg, S. W. Boettcher, J. A. Petykiewicz,
D. B. Turner-Evans, M. C. Putnam, E. L. Warren,
J. M. Spurgeon, R. M. Briggs, N. S. Lewis and
H. A. Atwater, Nat. Mater., 2010, 9, 239244.
253 J. M. Spurgeon, S. W. Boettcher, M. D. Kelzenberg,
B. S. Brunschwig, H. A. Atwater and N. S. Lewis, Adv.
Mater., 2010, 22, 32773281.
254 J. R. McKone, E. L. Warren, M. J. Bierman, S. W. Boettcher,
B. S. Brunschwig, N. S. Lewis and H. B. Gray, Energy
Environ. Sci., 2011, 4, 35733583.
255 Y. J. Hwang, C. H. Wu, C. Hahn, H. E. Jeong and
P. D. Yang, Nano Lett., 2012, 12, 16781682.
256 J. W. Sun, C. Liu and P. D. Yang, J. Am. Chem. Soc., 2011,
133, 1930619309.
257 X. C. Wang, K. Maeda, A. Thomas, K. Takanabe, G. Xin,
J. M. Carlsson, K. Domen and M. Antonietti, Nat. Mater.,
2009, 8, 7680.
258 Y. Zheng, J. Liu, J. Liang, M. Jaroniec and S. Z. Qiao, Energy
Environ. Sci., 2012, 5, 67176731.
259 G. Liu, P. Niu, C. Sun, S. C. Smith, Z. Chen, G. Q. Lu and
H.-M. Cheng, J. Am. Chem. Soc., 2010, 132, 1164211648.
260 J. S. Zhang, X. F. Chen, K. Takanabe, K. Maeda, K. Domen,
J. D. Epping, X. Z. Fu, M. Antonietti and X. C. Wang, Angew.
Chem., Int. Ed., 2010, 49, 441444.
This journal is c The Royal Society of Chemistry 2013
View Article Online
Chem Soc Rev
261 T. Dittrich, S. Fiechter and A. Thomas, Appl. Phys. Lett.,
2011, 99, 084105.
262 X. C. Wang, K. Maeda, X. F. Chen, K. Takanabe, K. Domen,
Y. D. Hou, X. Z. Fu and M. Antonietti, J. Am. Chem. Soc.,
2009, 131, 16801681.
263 J. S. Zhang, M. Grzelczak, Y. D. Hou, K. Maeda, K. Domen,
X. Z. Fu, M. Antonietti and X. C. Wang, Chem. Sci., 2012, 3,
443446.
264 A. Henglein, Chem. Rev., 1989, 89, 18611873.
265 S. Trasatti, J. Electroanal. Chem., 1980, 111, 125131.
266 S. Trasatti, J. Electroanal. Chem., 1972, 39, 163184.
267 K. Kalyanasundaram and M. Gratzel, Angew. Chem., Int.
Ed., 1979, 18, 701702.
268 J. M. Lehn, J. P. Sauvage and R. Ziessel, New J. Chem., 1979,
3, 423427.
269 J. M. Lehn, J. P. Sauvage and R. Ziessel, New J. Chem., 1980,
4, 623627.
270 J. M. Lehn, J. P. Sauvage, R. Ziessel and L. Hilaire, Isr. J.
Chem., 1982, 22, 168172.
271 V. Subramanian, E. E. Wolf and P. V. Kamat, J. Am. Chem.
Soc., 2004, 126, 49434950.
272 A. Takai and P. V. Kamat, ACS Nano, 2011, 5, 73697376.
273 B. H. Meekins and P. V. Kamat, J. Phys. Chem. Lett., 2011, 2,
23042310.
274 J. Kiwi and M. Gratzel, J. Am. Chem. Soc., 1979, 101, 72147217.
275 J. Kiwi and M. Gratzel, Nature, 1979, 281, 657658.
276 J. Kiwi and M. Gratzel, Angew. Chem., Int. Ed., 1979, 18,
624626.
277 A. Mills, J. Chem. Soc., Chem. Commun., 1982, 367368.
278 L. L. Tinker, N. D. McDaniel, P. N. Curtin, C. K. Smith,
M. J. Ireland and S. Bernhard, Chem.Eur. J., 2007, 13,
87268732.
279 B. F. DiSalle and S. Bernhard, J. Am. Chem. Soc., 2011, 133,
1181911821.
280 P. N. Curtin, L. L. Tinker, C. M. Burgess, E. D. Cline and
S. Bernhard, Inorg. Chem., 2009, 48, 1049810506.
281 A. Iwase, H. Kato and A. Kudo, Catal. Lett., 2006, 108, 69.
282 K. Maeda, A. K. Xiong, T. Yoshinaga, T. Ikeda,
N. Sakamoto, T. Hisatomi, M. Takashima, D. L. Lu,
M. Kanehara, T. Setoyama, T. Teranishi and K. Domen,
Angew. Chem., Int. Ed., 2010, 49, 40964099.
283 K. Maeda, K. Teramura, D. L. Lu, N. Saito, Y. Inoue and
K. Domen, J. Phys. Chem. C, 2007, 111, 75547560.
284 A. Sartorel, M. Carraro, G. Scorrano, R. De Zorzi,
S. Geremia, N. D. McDaniel, S. Bernhard and
M. Bonchio, J. Am. Chem. Soc., 2008, 130, 50065007.
285 Q. S. Yin, J. M. Tan, C. Besson, Y. V. Geletii, D. G. Musaev,
A. E. Kuznetsov, Z. Luo, K. I. Hardcastle and C. L. Hill,
Science, 2010, 328, 342345.
286 K. Maeda, K. Teramura, D. L. Lu, N. Saito, Y. Inoue and
K. Domen, Angew. Chem., Int. Ed., 2006, 45, 78067809.
287 Y. X. Zhao, N. M. Vargas-Barbosa, E. A. Hernandez-Pagan
and T. E. Mallouk, Small, 2011, 7, 20872093.
288 Y. X. Zhao, E. A. Hernandez-Pagan, N. M. Vargas-Barbosa,
J. L. Dysart and T. E. Mallouk, J. Phys. Chem. Lett., 2011, 2,
402406.
Chem. Soc. Rev., 2013, 42, 2294--2320 2319

Chem Soc Rev
289 M. Hara, C. C. Waraksa, J. T. Lean, B. A. Lewis and
T. E. Mallouk, J. Phys. Chem. A, 2000, 104, 52755280.
290 M. Hara and T. E. Mallouk, Chem. Commun., 2000, 19031904.
291 W. J. Youngblood, S. H. A. Lee, Y. Kobayashi,
E. A. Hernandez-Pagan, P. G. Hoertz, T. A. Moore,
A. L. Moore, D. Gust and T. E. Mallouk, J. Am. Chem.
Soc., 2009, 131, 926927.
292 H. Hata, Y. Kobayashi, V. Bojan, W. J. Youngblood and
T. E. Mallouk, Nano Lett., 2008, 8, 794799.
293 S. H. A. Lee, Y. X. Zhao, E. A. Hernandez-Pagan, L. Blasdel,
W. J. Youngblood and T. E. Mallouk, Faraday Discuss.,
Published on 16 October 2012. Downloaded by VIT University on 31/03/2017 11:01:47.
2012, 155, 165176.
294 K. Domen, S. Naito, M. Soma, T. Onishi and K. Tamaru,
J. Chem. Soc., Chem. Commun., 1980, 543544.
295 K. Domen, S. Naito, T. Onishi and K. Tamaru, J. Phys.
Chem., 1982, 86, 36573661.
296 T. Takata, Y. Furumi, K. Shinohara, A. Tanaka, M. Hara,
J. N. Kondo and K. Domen, Chem. Mater., 1997, 9,
10631064.
297 S. Ikeda, M. Hara, J. N. Kondo, K. Domen, H. Takahashi,
T. Okubo and M. Kakihana, Chem. Mater., 1998, 10, 7277.
298 A. Kudo, A. Tanaka, K. Domen, K. Maruya, K. Aika and
T. Onishi, J. Catal., 1988, 111, 6776.
299 A. Kudo, H. Kato and S. Nakagawa, J. Phys. Chem. B, 2000,
104, 571575.
300 K. Sayama, A. Tanaka, K. Domen, K. Maruya and T. Onishi,
J. Catal., 1990, 124, 541547.
301 K. Domen, A. Kudo and T. Onishi, J. Catal., 1986, 102, 9298.
302 Y. Gorlin and T. F. Jaramillo, J. Am. Chem. Soc., 2010, 132,
1361213614.
303 F. Jiao and H. Frei, Angew. Chem., Int. Ed., 2009, 48,
18411844.
2320 Chem. Soc. Rev., 2013, 42, 2294--2320
View Article Online
Review Article
304 R. Nakamura and H. Frei, J. Am. Chem. Soc., 2006, 128,
1066810669.
305 M. W. Kanan and D. G. Nocera, Science, 2008, 321,
10721075.
306 N. H. Chou, P. N. Ross, A. T. Bell and T. D. Tilley, Chem-
SusChem, 2011, 4, 15661569.
307 D. Shevchenko, M. F. Anderlund, A. Thapper and
S. Styring, Energy Environ. Sci., 2011, 4, 12841287.
308 T. F. Jaramillo, K. P. Jorgensen, J. Bonde, J. H. Nielsen,
S. Horch and I. Chorkendor, Science, 2007, 317, 100102.
309 J. Bonde, P. G. Moses, T. F. Jaramillo, J. K. Norskov and
I. Chorkendor, Faraday Discuss., 2008, 140, 219231.
310 Y. Hou, B. L. Abrams, P. C. K. Vesborg, M. r. E. Bjorketun,
K. Herbst, L. Bech, A. M. Setti, C. D. Damsgaard,
T. Pedersen, O. Hansen, J. Rossmeisl, S. r. Dahl,
J. K. Nrskov and I. Chorkendor, Nat. Mater., 2011, 10,
434438.
311 H. A. Atwater and A. Polman, Nat. Mater., 2010, 9, 205213.
312 A. Valdes, J. Brillet, M. Gratzel, H. Gudmundsdottir,
H. A. Hansen, H. Jonsson, P. Klupfel, G. J. Kroes, F. Le
Formal, I. C. Man, R. S. Martins, J. K. Norskov,
J. Rossmeisl, K. Sivula, A. Vojvodic and M. Zach, Phys.
Chem. Chem. Phys., 2012, 14, 4970.
313 H. W. Gao, C. Liu, H. E. Jeong and P. D. Yang, ACS Nano,
2012, 6, 234240.
314 E. Thimsen, F. Le Formal, M. Gratzel and S. C. Warren,
Nano Lett., 2011, 11, 3543.
315 Z. B. Chen, T. F. Jaramillo, T. G. Deutsch, A. Kleiman-
Shwarsctein, A. J. Forman, N. Gaillard, R. Garland,
K. Takanabe, C. Heske, M. Sunkara, E. W. McFarland,
K. Domen, E. L. Miller, J. A. Turner and H. N. Dinh,
J. Mater. Res., 2010, 25, 316.
This journal is c The Royal Society of Chemistry 2013